(navigation image)
Home American Libraries | Canadian Libraries | Universal Library | Community Texts | Project Gutenberg | Children's Library | Biodiversity Heritage Library | Additional Collections
Search: Advanced Search
Anonymous User (login or join us)
Upload
See other formats

Full text of "Bulletin of the British Museum (Natural History) Geology Supplement"

IBfti 









UPPER JURASSIC ^/ 

RHYNCHONELLID BRACHIOPODS 
FROM NORTHWESTERN EUROPE 



A. CHILD S 



BULLETIN OF 

THE BRITISH MUSEUM (NATURAL HISTORY) 

GEOLOGY Supplement 6 

LONDON: 1969 



UPPER JURASSIC RHYNCHONELLID 
BRACHIOPODS FROM NORTHWESTERN 

EUROPE 



BY 




ALAN CHILDS 

(College of Technology, Oxford) 



12 plates ; 40 text- figures 



BULLETIN OF 
THE BRITISH MUSEUM (NATURAL HISTORY) 
GEOLOGY Supplement 6 

LONDON: 1969 



THE BULLETIN OF THE BRITISH MUSEUM 

(natural history), instituted in 1949, is 
issued in five series corresponding to the Departments 
of the Museum, and an Historical series. 

Parts will appear at irregular intervals as they 
become ready. Volumes will contain about three or 
four hundred pages, and will not necessarily be com- 
pleted within one calendar year. 

In 1965 a separate supplementary series of longer 
papers was instituted, numbered serially for each 
Department. 

This paper is Supplement No. 6 of the Geological 
(Palaeontological) series. The abbreviated titles of 
periodicals cited follow those of the World List of 
Scientific Periodicals. 



World List abbreviation 
Bull. Br. Mus. nat. Hist. (Geol.) Suppl. 



Trustees of the British Museum (Natural History) 1969 



TRUSTEES OF 
THE BRITISH MUSEUM (NATURAL HISTORY) 

Issued 19 September, 1969 Price £4 15s. 



UPPER JURASSIC 

RHYNCHONELLID BRACHIOPODS FROM 

NORTHWESTERN EUROPE 

By ALAN CHILDS 



CONTENTS 

I. Introduction ...... 

Previous Research ..... 

II. Ecology ....... 

III. Geographical and Stratigraphical Distribution 

IV. Morphology ...... 

V. Systematic Descriptions 

Genus Monticlarella. .... 

Lacunosella ..... 
Rhynchonella . .... 

Thurmannella .... 

Acanthothiris .... 

Acanthorhynchia .... 
Somalirhynchia .... 
Septaliphoria . .... 

Torquirhynchia .... 
" Rhynchonella " . 
VI. Acknowledgments ..... 
VII. References. ...... 



Page 
3 
4 
6 

13 
15 

19 

28 

41 

45 
53 
63 
78 
85 
95 
107 

"3 
114 



SYNOPSIS 

The major part of this study comprises a systematic revision of most of the Upper Jurassic 
rhynchonellids of northwestern Europe. Thirty-five species belonging to the genera Monti- 
clarella, Lacunosella, Rhynchonella, Thurmannella, Acanthothiris, Acanthorhynchia, Somali- 
rhynchia, Septaliphoria, and Torquirhynchia gen. nov. are described. By using serial sectioning 
and acetate peel techniques, the internal structures of all but the rarest species have been in- 
vestigated and the majority of these are described and figured for the first time. New taxa in- 
clude six species, a subgenus, and a genus. 

Observations on ecology and distribution are given with the specific descriptions and separate 
sections are also included dealing with these aspects at generic level. 



INTRODUCTION 

In contrast to the many monographs describing the complete faunas of small areas, 
this study consists of the examination of a single order, the Rhynchonellida, over a 
wider region ; the author considering that this approach provides a much better 
basis for taxonomic as well as ecological and distributional studies. The rhyn- 
chonellid brachiopods are a morphologically diverse group and, although not 
numerous in the Upper Jurassic of the British Isles, they colonised a wide range of 
habitats in N.W. Europe during this period. 



4 JURASSIC RHYNCHONELLIDS 

Although most of the species described have been previously figured, this is the 
first attempt to give such a comprehensive account of these faunas and to clarify their 
synonymies. It is interesting to note in this connection that the French literature 
abounds with Orbigny and Lamarck species, the German with those of Schlotheim, 
Oppel and Quenstedt while many British authors have relied on Sowerby and 
Davidson. By making a palaeontological study on a wider geographical basis than 
is usually attempted, it is hoped to reduce the number of species confined by national 
boundaries. It is considered that the other major justification for attempting to 
study a group over a relatively wide geographical area is that in considering evolu- 
tionary lineages this helps to eliminate any local distortion caused by migration and 
the general imperfection of the fossil record. 

Although essentially work has been confined to taxa within the Upper Jurassic, 
certain species and genera from lower stratigraphical levels have been investigated 
where this was thought necessary. The type species of Acanthorhynchia, namely 
A. panacanathina, from the Bajocian has thus been included and similarly a study 
was made of Acanthothiris spinosa in order to determine the justification for splitting 
the spinose forms into separate genera. 

Previous research 

Although no previous author has attempted to produce such a comprehensive 
monograph of this group of rhynchonellids, there have been numerous descriptions 
and figures published over the last two hundred years. As there are relatively few 
rhynchonellids in the British Upper Jurassic rocks, most of the works referred to are 
by continental authors. 

Linnaeus (1767) described only one of the species discussed below, namely Acan- 
thothiris spinosa ; this species was also the subject of the first figure to which 
reference is made, Knorr and Walch (1769). This appears to be the only recognisable 
eighteenth century description of a Middle or Upper Jurassic rhynchonellid. Of the 
workers of the early part of the nineteenth century, the most important were 
Schlotheim (1813, 1820), Zieten (1830-33) and, to a lesser extent, Lamarck (1819). 
The paucity of British Upper Jurassic material meant that the Sowerbys figured 
only one species — Torquirhynchia inconstans. Fischer de Waldheim's paper of 1809 
must also be mentioned as this introduced the genus Rhynchonella. The major 
nineteenth century French work was undoubtedly d'Orbigny's (1850) "Prodrome ", 
but this lost much of its potential value through lack of illustrations. 

From about 1850 until the beginning of the present century several eminent 
palaeontologists produced many of the monographs still regarded as standard works. 
These writers included Davidson, who described and figured all the then known 
British brachiopods, Eudes-Deslongchamps, who produced numerous well illustrated 
papers on the brachiopod faunas of Normandy and elsewhere and Loriol and Haas 
who produced similar comprehensive works on the French and Swiss Jura. Many 
of the monographs of these authors set standards which have seldom been surpassed. 
Mention must also be made of the writings of Quenstedt (1851-52), which culminated 
in his magnificent work " Die Brachiopoden " (1868-71), still the most complete 



FROM NORTHWESTERN EUROPE 5 

illustrated account of European brachiopods. Apart from these, there are also 
numerous papers and monographs dealing with smaller areas such as those of 
Struckmann (1878) on the region of Hannover and Rothpletz (1886) on the Vilser 
Alps. The latter, while strictly outside the area under consideration, includes 
important discussion on the higher taxonomy of many of the species described. 



UPPER JURASSIC 


ZONES AND STAGES 


nodiger 


UPPER VOLGIAN 


subditus 


fulgens 


giganteus 


LOWER VOLGIAN 


gorei 


albani 


pallasioides 


rotunda 


pectinatus 


wheatleyensis 


Subplanites spp. 


Gravesia spp. 


pseudomutabilis 


KIMMERIDGIAN 


mutabilis 


cymodoce 


baylei 


pseudocordata 


OXFORDIAN 


transversarium 


cordatum 


mariae 



Fig. 1. The nomenclature of the Upper Jurassic zones and stages used in this study, following 
the recommendation of the British Mesozoic Committee as reported by Ager (1964). 

During the present century there has been relatively little investigation of the 
Upper Jurassic rhynchonellids. Buckman (1918), as part of his " Burma Memoir ", 
completely revised the taxonomy of the Jurassic rhynchonellids but this revision 
was largely based on Lower and Middle Jurassic British material and few of the 
species here described can be attributed to his genera. With regard to Buckman's 
work, Ager (1956) wrote, " His (Buckman's) classification is not now accepted, his 
terminology is unnecessarily complex and his emphasis in description is frequently 
misdirected ". In 1917 Rollier produced his " Synopses " in which he attempted to 
elucidate the synonymies and redefine all Jurassic brachiopod species. While this 
work contains many useful suggestions, it was apparently undertaken without any 
reference to original specimens and consequently contains many misconceptions as a 



6 JURASSIC RHYNCHONELLIDS 

result of referring to figures and descriptions which were inadequate in the first 
place. In 1920 Leidhold published a paper in which he proposed the genus Septali- 
phoria and two subgenera within it, namely Thurmannella and Blochmannella. 
Although the descriptions of the type species were inadequate, the names Septali- 
phoria and Thurmannella have since been widely used as a result of the lack of 
generic names at this level. 

Probably the most notable contribution to the subject was Wisniewska's (1932) 
monograph on the Upper Jurassic rhynchonellids of Poland. In that work she pro- 
posed the genera Monticlarella, Lacunosella and Septocrurella and in publishing the 
first transverse serial sections of many of the rhynchonellids of this age made a major 
advance in their study. The use of transverse serial sections was well known to 
several authors of the last century, notably Bittner and Rothpletz, but the tech- 
nique was allowed to lapse until its use by Wisniewska and its application was not 
widely publicised until Muir- Wood's paper in 1934. In common with virtually all 
the other works previously mentioned, Wisniewska gave no details of lithology or 
associated fauna and very little information about relative abundance and distribu- 
tion. 

Since Wisniewska, the only major work devoted to the Upper Jurassic brachiopods 
has been Makridin's (1964) study of the faunas of the Russian Platform. In this 
work Makridin was seriously handicapped by a lack of comparative material from 
the classic areas of N.W. Europe and, as a result, many of his assignations and con- 
clusions are considered to be doubtful. Brief diagnoses of all the Mesozoic genera 
are to be found in Makridin (i960) and Ager (1965&). The ecological aspects of the 
Mesozoic brachiopod faunas have recently been discussed by Makridin (1964) and 
Ager (1965a). 

II. ECOLOGY 

The environments colonised by the rhynchonellids during the whole of the Meso- 
zoic were discussed in detail by Ager (1965a). In that paper seven different habitats 
were recognized namely : (1) very shallow water sea floors, (2) sublittoral, sand 
grade sea floors without reefs, (3) sea floors in the vicinity of reefs, (4) shallow, non- 
depositional sea floors, (5) sublittoral, mud-grade sea floors, (6) deeper (? bathyal) 
mud-grade sea floors, (7) floating weed, and the adaptations shown by the Mesozoic 
brachiopods to each of them described. Within the area described by the present 
author, however, not all these environments appear to be present as the region was 
largely covered by a relatively shallow shelf sea. As far as they have been eluci- 
dated, the environments colonised by the rhynchonellids during the Upper Jurassic 
in northwestern Europe are as follows : 
(a) Sandy, near-shore environments. 

This facies is well developed in the lower and middle Oxfordian around the northern 
rim of the Paris Basin ; a similar environment was also present in much the same 
area during lower Callovian times and, while this is strictly outside the strati- 
graphical limits of this study, it is discussed for comparison. The formations 
representing this environment at these times were the " Terrain a Chailles " and 



FROM NORTHWESTERN EUROPE 7 

the " Varians Schichten " respectively. Both of them are sandy limestones, 
although they differ in that the former contains nodular layers of chert. Each of 
these formations contains an abundance of rhynchonellids and in each of them two 
markedly different types are present. The two rhynchonellids occurring in the 

' Terrain a Chailles " are Septaliphoria arduennensis (Oppel) and Thurmannella 
obtrita (Defrance) . The former is much the larger and in having coarse ribs, a strong 
beak and a large pedicle opening was obviously adapted to a high energy environ- 
ment. The much more abundant T. obtrita has relatively much finer ribbing and a 
very thin test, although it too has a well developed pedicle opening. Within the 

' Varians Schichten " the rhynchonellids are of much the same size and occur in 
similar proportions. In this case the larger is Acanthothiris spinosa (Linnaeus) and 
the smaller species is Ivanoviella alemanica (Rollier) . The former, like S. arduennen- 
sis, tends to occur singly rather than the large groups in which T. obtrita and Ivano- 
viella alemanica are found. The latter species closely resembles T. obtrita externally 
except that it has stronger ribs and a more limited posterior smooth area. It is 
suggested that the similarity of environments in which A . spinosa and S. arduennensis 
are found, and the fact that the latter species is obviously adapted to living an 
attached existence in such conditions is strong evidence for regarding the spines of 
A. spinosa as an anchoring mechanism, especially as its pedicle atrophied during 
ontogeny. This view is in contrast to that of Rudwick (1965) who considered these 
spines to be of a sensory nature. It is thought that the smaller, thinner shelled T. 
obtrita and /. alemanica probably lived attached to algae. This idea is supported by 
the presence of the relatively large pedicle opening, while the very thin shell of T. 
obtrita, as seen in silicified material, would appear to preclude its survival on the 
bottom in such numbers in a high energy environment. However, it is always 
possible that they lived in sheltered niches within this environment. 

The same type of environment is suggested for the deposition of the Lower 
Calcareous Grit as seen on the Yorkshire coast. In that area Thurmannella is again 
abundant but any larger rhynchonellids appear to be very much rarer although 
Mr. E. F. Owen (personal communication, 1966) states that he has found Septali- 
phoria on Filey Brigg. The material described below as S. paucicosta sp. nov. came 
from inland localities but it is not known whether or not the species is associated 
with Thurmannella sp. The amount of material in the B.M. (N.H.) would suggest 
that S. paucicosta is very common at least at its type locality of Hutton Bushel. 

The nearest comparable lithology and environment in the Kimmeridgian is pro- 
vided by the Alt-na-Cuile Sandstone of Sutherland. This latter formation is a 
decalcified sandstone containing a rich fauna of rhynchonellids preserved as casts. 
As far as can be determined, given the poor preservation, this seems to be a coarse 
ribbed form closely comparable with Septaliphoria but as its internal structures are 
not preserved its exact relationship cannot be determined. This latter species is 
described below under the name Septaliphoria (?) septentrionalis sp. nov. The 
Abbotsbury Ironstone (Kimmeridgian) of Dorset may also represent this type of 
environment. Here again, the fauna includes poorly preserved rhynchonellids 
doubtfully ascribed to Septaliphoria (?) hudlestoni (Rollier). 



8 JURASSIC RHYNCHONELLIDS 

Ager (1965a) suggested that the strongly uniplicate rhynchonellids such as 
Homoeorhynchia acuta and Rhynchonella loxiae were characteristic of sand grade, 
sublittoral sea floors. However, it is not thought that all species of the genus 
Rhynchonella lived in this type of environment as at least two of them, namely, 
R. subvariabilis and R. rivelensis, occur in clays or marls. It appears from the 
relatively limited information available that the earlier species attributable to 
Rhynchonella s.s. are always found in such lithologies and that it is not until Volgian 
times that the genus is found in a more sandy facies which presumably represents a 
higher energy environment. Makridin (1964), who gives the range of Rhynchonella 
as late Kimmeridgian — early Cretaceous, states that, " members of the genus are 
most widespread in the shoaly facies of a sandy and sandy-clay sublittoral ". If the 
above information is correct, then it is interesting to note that the genus became 
much more morphologically diverse on colonising the higher energy environment. 
It is possible that Rhynchonella only colonised these environments on the extinction 
of such genera as Septaliphoria. In order to account for the occurrence of R. 
subvariabilis, Ager suggested a pelagic mode of life and it is thought that such a 
mode of life would also best account for the occurrences of the earlier species of the 
genus such as R. rivelensis and R. triplicosa, this is discussed further below. 
(b) Sea floors in the vicinity of reefs. 

While Ager (1965a) referred exclusively to the forms encountered in the vicinity 
of hermatypic coral reefs, within the Upper Jurassic, it is possible to subdivide this 
environment according to whether the reefs are coral or sponge. The greatest 
development of sponge reefs is found in the regions of Swabia and Franconia where 
they flourished from middle Oxfordian to, at least, lower Volgian times. The 
rhynchonellid fauna associated with these reefs is very distinctive and consists almost 
entirely of species of the genus Lacunosella. The incidence of sponges with Lacuno- 
sella spp. is such as to suggest that the rhynchonellids were actually dependent on 
them. Middlemiss (1962) has recorded rhynchonellids within the folds of Rhaphi- 
donema in the Lower Greensand at Faringdon, Berkshire, and this may possibly be a 
commensal relationship ; unfortunately, details of the rhynchonellids were not 
given. It has also been suggested (Ager, 1965a) that Orbirhynchia of the Upper 
Cretaceous adopted a similar mode of life and it may be significant that both the 
latter genus and Lacunosella possess the distinctive blade-like falcifer crura. Apart 
from Lacunosella, the only other rhynchonellid genera known to occur in the sponge 
reef facies are Acanthorhynchia (Echinirhynchia) and Monticlarella ; these may well 
have been pelagic forms and are discussed further below. 

The important factor in determining the distribution of the lacunosellids would 
seem to be the presence of sponges and whether these were living as reef assemblages, 
loose groups or " lenses " mattered little. While the greatest abundance and 
variety of lacunosellids is found in the sponge reefs of Germany where sponges were 
the dominant feature of the fauna during the Upper Jurassic, in the southern French 
Jura L. arolica occurs in calcareous shales with only isolated sponges. 

The areas around the coral reefs appear to have been colonised by two genera. 
The more distinctive externally is Torquirhynchia gen. nov., which comprises a 



FROM NORTHWESTERN EUROPE 9 

series of strongly asymmetrical species, as typified by T. inconstans (J. Sowerby) 
from the Kimmeridge Clay. The reason for this somewhat bizarre development is 
not understood ; however, it is interesting to note that the individuals concerned 
can be inverted and still retain the same form of anterior commissure. This is 
particularly obvious in the more globose specimens of T. inconstans in which the 
brachial valve was inflated to such an extent that the pedicle must have atrophied. 
Given the subspherical appearance of these inflated specimens, their lack of a pedicle 
and the fact that they apparently lived in a high energy, perireefal environment, it 
seems reasonable to assume that the development of asymmetry was related to these 
facts. That Torquirhynchia species can be inverted and still retain the same form of 
anterior commissure may possibly have been of advantage in that, as the right and 
left halves of the mantle cavity are physiologically independent (Orton, 1914), at 
least half of the lophophore system would have a water intake well clear of the 
bottom. 

With the exception of T. cf. T. astieriformis, which was collected from around a 
small sponge reef, the genus has always been found associated either with corals or 
with " reef " limestones. Torquirhynchia seems to be completely absent from the 
sponge reef facies of the Swabian and Franconian Alb ; T. speciosa is found in the 
Kelheim area but the " Diceras Kalk " in which it occurs has much closer affinities 
with the fauna of the Stramberk Limestone of Czechoslovakia. 

Apart from Torquirhynchia, the only other genus to have colonised the areas 
around coral reefs is Somalirhynchia Weir. 5. moeschi (Haas), which occurs widely 
in the central French Jura, is always associated with corals according to M. Enay 
(personal communication, 1965) and this was certainly so at the one locality where 
it was collected by the author. It is considered that the presence of corals in the 
Kimmeridgian Boulder Beds of Sutherland is significant in accounting for the 
occurrence of S. sutherlandi. This occurrence of S. sutherlandi is discussed in some 
detail under the specific description. It must be stressed that both Torquirhynchia 
and Somalirhynchia appear to be strictly perireefal. Elliot (1950) has suggested that 
the absence of brachiopods within reefs is probably a result of the coral polyps eating 
the brachiopod larvae. 

(c) Sublittoral, muddy sea floors. 

This is the remaining major environment in which rhynchonellids have been 
collected. As stated by Ager, it seems reasonable to assume that this is a deeper 
water facies than those discussed above, but this need not be so and the present 
writer would agree that the grain size of the substratum seems to be more important 
than the actual depth. With regard to this latter point, at various localities in the 
southern French Jura Lacunosella arolica occurs in a marly facies. However, as 
stated above, it is invariably associated with sponges and it is considered that these 
represented the substrate as far as the lacunosellids were concerned and consequently 
the presence of sponges is the determining factor rather than sediment or depth. 

Other genera which occur in this environment include Rhynchonella, Thurmannella, 
Echinirhynchia and Monticlarella. Of these it is suggested that forms such as T. 



JURASSIC RHYNCHONELLIDS 



ohtrita from the " Renggcri Marl " and R. rivelensis and R. subvariabilis from the 
upper Oxfordian and upper Kimmeridgian respectively were probably attached to 
algae during life. This is suggested as the normal mode of existence but does not 
preclude occasional specimens being drifted considerable distances and thus account- 
ing for the occurrence of rare species such as subvariabilis, as suggested by Ager. 
T. obtrita and R. rivelensis occur in such abundance that it seems unlikely that they 




Fig. 2. The solid black area represents the outcrop of the Callovian-Volgian ; this informa- 
tion is taken from the map commissioned by the World Geological Congress of 1881. A 
generalized palaeogeography of the Oxfordian-Kimmeridgian (after Kuhn, 1953) is shown 
with the land areas ornamented with crosses. 



FROM NORTHWESTERN EUROPE n 

are not in the area where they lived and this is borne out by the fact that onto- 
genetic series of T. obtrita can be collected. 

Concerning the remaining genus and subgenus, namely Monticlarella and Echi- 
nirhynchia, the latter is known to occur in all the facies previously described apart 
from in association with corals while Monticlarella is recorded from all except the 
near shore, high energy type. These observations, coupled with the small size of the 




Distribution of Oxfordian 
rhynchonellid genera 



Th - B 

Rh La° 
So La 



ip 




Fig. 3. Ac = Acanthorhynchia ; La= Lacunosella ; Rh = Rhynchonella ; " Rh ' 
chonella " ; Se = Septaliphoria ; So = Somalirhynchia ; Th = Thurmannella. 



Rhyn- 



12 JURASSIC KHYNCHONELL1DS 

taxa concerned, suggest that they were attached to algae or sponges and consequently 
easily drifted long distances. While the normal mode of life may well have been the 
same as that of Lacunosella, the small size and delicate nature of the test obviously 
made it much easier for them to be transported over considerable distances and 
account for their appearance in such a diverse range of habitats. 




Distribution of Kimmeridgian 
rhynchonellid genera 



Mo 



"Rh" 


e« Mo 


•N 


La 

r. La Mo 

, Mo 

La m- 


To 




Ac 
l- To Mo 





•p 




Fig. 4. Ac = A canthorhynchia ; La = Lacunosella ; Mo = M onticlarella ; Kh = Rhynchonella ; 
" Rh " = " Rhynchonella "; Se?=Seplaliphoria? ; To = Torquirhynchia. 



FROM NORTHWESTERN EUROPE 
GEOGRAPHICAL AND STRATIGRAPHICAL DISTRIBUTION 



13 



Details of the distribution of individual species are given in the systematic part of 
this study but text-figs. 3-5 have been included to summarise the distribution of the 
genera. These figures are based on information obtained from field observations, 
museum collections and literature. It is considered that the distribution is largely 
controlled by fades and substrate, as shown by the occurrence of Septaliphoria and 
Lacunosella. 



Distribution of Lower Volgian 
rhynchonellid genera 




t- . Mo 
La 



La 

E . Ac 
Mo 

La L ° To 




Fig. 5. Ac = A canthorhynchia ; La = Lacunosella ; Mo = Monticlarella ; Rh = Rhynchonella 

To = Torquirhynchia . 



14 



JURASSIC RHYNCHONELLIDS 



Oxfordian 



Kimmeridgian 



Lower Volgian 



M czenstochaviensis 



M striocmcta 



MONTICLARELLA 



M. triloboides 
M. strioplicata 



— » L. vaga 

L. eracoviensis 
L. sparsieosta 
— -^— L trilobata 



LACUNOSELLA 



R. portlandica 
R. subvanabilis — 



R rivelensis 



RHYNCHONELLA 



R. sp from Sutherland 



T oblrita 
T. acuticosta 



THURMANNELLA 



A spinulosa 
A lorioli 



ACANTHORHYNCHIA 



A senticosa 



A. fileyensis 



S. Sutherland! 



SOMALIRHYNCHIA 



S moeschi 



S. arduennensis 
S pauacosta 

S ? septentrionalis 

— — S ' hudlesloni 



SEPTALIPHOR/A 



T inconstans 
• T guebhardi 

^^— T aslieriformis 



TORQUIRHYNCHIA 



J. speciosa 



— R ordinana 

R pyrennaei 



"RHYNCHONELLA" 



Fig. 6. The stratigraphical ranges of the Upper Jurassic rhynchonellids here described. 



FROM NORTHWESTERN EUROPE 15 

In order to help clarify the distribution patterns, text-fig. 2 shows the outcrop of 
the Upper Jurassic (including the Callovian) and also indicates a generalised palaeo- 
geography of Oxfordian-Kimmeridgian times, after Kuhn (1953). Volgian times 
saw a major withdrawal of the sea from much of the area and the consequent marked 
decrease in the rhynchonellid faunas can be seen by comparing text-figs. 4 and 5. 
This decrease is particularly evident when it is considered that most of the German 
fauna plotted on text-fig. 5 is probably restricted to the Gravesia and Subplanites 
zones of the basal Volgian. 

The stratigraphical ranges of the Oxfordian- Volgian species studied are tabulated 
on text-fig. 6. These ranges have been determined as accurately as possible, but it 
has often proved difficult to translate records from older stratigraphical papers into 
modern zonal terms. 

MORPHOLOGY 

The present author has accepted the definitions of morphological terms used in 
the " Treatise " (1965) with the exception of the following points : 

Within the species studied, it has been found when describing the form of the 
pedicle opening the choice of term has invariably been limited to either hypo- 
thyridid or submesothyridid. In many cases it is extremely difficult to differentiate 
these types ; in none of the species examined did the pedicle opening strongly 
impinge on the beak ridges. The second point concerns the pedicle collar ; accord- 
ing to the definition in the " Treatise ", this is an internal feature of the ventral 
beak, " continuous laterally with the internal surface of the deltidial plates ". 
However, the feature figured by Thomson (1927) under that name consisted of a 
groove running from the posterior side of the pedicle opening to the tip of the beak. 
This latter structure has been found in the genus Acanthothiris and the term 
" pedicle trough " is proposed for it. 

In describing the internal structures, care has been taken to stress when the 
features are being described as seen in transverse section and also to avoid mis- 
leading descriptions such as " septum long " when in fact it is high. The difficulties 
present in interpreting transverse sections are further discussed below with reference 
to the septalium. 

The septalium has recently been defined in the " Treatise " as being a " Trough- 
like structure of the brachial valve between the hinge plates (or homologues), con- 
sisting of septalial plates (or homologues), enveloping and buttressed by median 
septum ; does not carry adductor muscles"; the author would agree with this 
definition. However, it has been thought necessary to discuss this structure in 
some detail for two main reasons ; firstly, the author has collected silicified material 
of Septaliphoria arduennensis and Thurmannella obtrita, the species in which this 
structure was originally described by Leidhold (1920), and secondly, because there 
seems to be some confusion in the literature as to the appearance of the septalium as 
seen in transverse section. 

All the elements which together constitute the septalium are capable of consider- 
able variation. This point would seem to be self-evident and yet has undoubtedly 



ii. JURASSIC RIIVNCHONELLIDS 

been the cause of some of the misunderstanding. The first full description of the 
septalium showing its appearance in transverse section was given by Muir-Wood 
(1934) when she chose as her example the zeilleriid Digonella digona (Sowerby), 
despite the fact that it had originally been defined with reference to a rhynchonellid. 
The septalium found in D. digona is relatively shallow and supported by a high 
septum. These features, coupled with the flattened nature of the brachial valve, 
result in the appearance, in transverse section, of a septalium bearing little resem- 
blance to the same structure as seen in most Mesozoic rhynchonellids. As stated 
above, the septalium in zeilleriids consists of a well developed, shallow trough lying 
between the hinge plates and supported for some distance by a high median septum. 
In contrast to this, the septalium in Septaliphoria arduennensis is only developed at 
the extreme posterior end of the valve and is either supported by a very low septum 
or, as pointed out by Wisniewska (1932), sometimes appears to rest directly on the 
floor of the valve. The other point, which was recently discussed by Rousselle 
(1965), is that the appearance of the septalium in transverse section largely depends 
on the degree of inflation shown by the brachial valve. 

Text-figure 7 shows the septalium developed in Septaliphoria arduennensis and 
also two transverse sections through the posterior part of the brachial valve of 5. 
paucicosta sp. nov. These demonstrate the way in which it is possible to have a 
septalium present and yet not to have the " diagnostic " U-shaped trough in the 
hinge plates developed in transverse section. From this it is evident that consider- 
able thought should be given to the way in which shell globosity, angle of sectioning 
and relative development can cause the same structure to show considerable varia- 
tion in appearance as seen in transverse section. Transverse sections of S. pauci- 
costa rather than S. arduennensis have been illustrated as the fine detail was better 
preserved in that species. Rousselle (1965) suggested describing the septalium as 
either " apparent " or " non apparent " depending on whether or not it appeared 
" trough shaped " in transverse section. This suggestion does not seem to be 
particularly useful as a septalium is either present or it is not and its appearance in 
transverse section is obviously dependent on the factors mentioned above. 

The transverse sections of S. paucicosta clearly show the way in which the septalial 
plates are united with the median septum. This latter point has been the subject of 
some discussion. The figured photomicrograph (pi. 12, fig. 6) of an acetate peel 
proves conclusively as has been pointed out by several authors (Ager, 19656), that 
Leidhold (1920) was mistaken in describing the septalium as, " arising from a 
bifurcation of the dorsal septum at its posterior end and fusion of the forked struc- 
tures with the hinge plate ". The definition in the " Treatise ", given above, is 
obviously the correct interpretation. A photomicrograph of a zeilleriid septalium 
is figured for comparison with that of S. paucicosta on pi. 12. 

In the genera described, the author has recognized four of the crural types so far 
defined namely, radulifer, calcarifer, arcuifer and falcifer. The author would agree 
with the " Treatise " definitions with the exception of one point concerning the 
radulifer type. In the material studied, the radulifer crural bases do not arise on the 
ventral side of the hinge plates but on the dorsal. This is demonstrated by photo- 



FROM NORTHWESTERN EUROPE 



hinge 
plate 



septatial 
plate 




socket 



median septum 



septatial 
plate 




median 

septum 



position of septalium (as 
seen in longitudinal section) 
in a rhynchonellid with 
an inflated brachial 
valve, (after Rousselle). 
with globose specimens 
the form of the septalium 
in transverse section is 
as figured above. 



sketch of septalium as 
seen in silicified speci- 
mens of Septaliphoria 
arduennensis (Oppel). 



transverse sections 
through posterior part 
of brachial valve of 
S.paucicosta sp. nov 



hinge plate 



septatial 
plate 

median 

septum 




Fig. 7. Sketches to show the form of the septalium in the genus Septaliphoria Leidhold. 



18 JURASSIC RHYNCHONELLIDS 

micrographs of an acetate peel of the crural bases of Septaliphoria paucicosta (pi. 12). 
In some species with radulifer crura such as Acanthothiris spinosa, the crural bases 
are only poorly differentiated. 

Calcarifer type crura have been found in the genera Thurmannella and Ivanoviella ; 
the form of the crura in these genera was strictly comparable to those observed in a 
specimen of Rhynchonelloidella smithi sectioned for comparison. Although, as 
pointed out in the " Treatise ", distally the calcarifer are similar in form to the 
falcifer, in mode of occurrence and proximal appearance they differ markedly. 
Text-figures showing the form of falcifer crura are given with the specific descriptions 
of Lacunosella arolica, L. sparsicosta, L. cracoviensis and L. vaga. The most striking 
difference is that, by using acetate peels, it can be seen that the early formed parts of 
the crura lie within the hinge plates of the species with calcarifer crura while in 
Lacunosella the hinge plates are never united and the early formed parts of the crura 
are external to them and never surrounded by hingeplate material (pi. 12). This 
feature also differentiates falcifer crura from the radulifer and arcuifer types, apart 
from considerations of gross morphology. The crural bases are formed by the 
deflection of hinge plate material around these early formed portions of the crura 
(pi. 12). Whether or not the crural bases are clearly differentiated depends partly 
on whether the early crus was rounded or sharply pointed, and also on how closely 
the hinge plate material followed the original shape. Although the early crus may 
have been sharply pointed, this shape may be lost completely or reduced to a 
rounded swelling on the dorsal side of the hinge plates where the latter are much 
thickened. 

Arcuifer crura are only recorded in the genus Monticlarella. Three species of that 
genus were investigated, using acetate peels, but the preservation was poor in all 
cases and no information additional to that given in the " Treatise " was obtained. 

Cardinal processes of the type found in certain Palaeozoic rhynchonellids such as 
the Uncinulidae are not developed in the material studied and, within the Upper 
Jurassic, only the genus Acanthorhynchia shows any comparable structure. In that 
genus the inner hinge plates are characteristically thickened and in some species 
this gives rise to a low rounded process on the ventral side of the hinge plates. A 
photomicrograph of the thickened hinge plates of Acanthorhynchia {Acanthorhynchia) 
panacanathina is figured on plate 12. 



Order RHYNCHONELLIDA Kuhn, 1949 

Superfamily RHYNCHONELLACEA Gray, 1848 

Family DIMERELLIDAE Buckman, 1918 

Subfamily MONTICLARELLINAE nov. 

Diagnosis. Small rhynchonellids ; ornament includes radial striae ; ribs 
absent or variably developed ; beak small and pointed ; slightly sulcate, recti- 
marginate or weakly uniplicate ; crura arcuifer where known. 

Stratigraphical range. Lower Jurassic — basal Cenomanian. 



FROM NORTHWESTERN EUROPE 19 

Genus MONTICLARELLA Wisniewska 

1932 Monticlarella Wisniewska, pp. 55-57. 

i960 Monticlarella Wisniewska; Makridin, p. 248. 

19656 Monticlarella Wisniewska; Ager, p. H604. 

Type species. Rhynchonella czenstochaviensis Roemer, by original designation. 
Emended diagnosis. Small subpentagonal or subtriangular rhynchonellids ; 
radial ornament always includes striae and usually ribs ; small, sharp beak; crura 
arcuifer. 

Stratigraphical range. Upper Jurassic — ? Lower Cretaceous. 
Description. External characters. All the included species are small, sub- 
pentagonal or subtriangular, symmetrical and biconvex. The valves may be of 
equal convexity or the pedicle valve may be the more inflated. The small, pointed, 
suberect to erect beak does not obscure the hypothyridid pedicle opening ; deltidial 
plates, when present, are small and disjunct. A small flattened interarea and 
distinct beak ridges are developed. 

One of the most distinctive features is the ornament which always includes radial 
striae, even though these may be restricted to the troughs between the ribs. Ribs, 
rounded or subrounded, are usually present, although not on the type species : the 
variation in the development of the ribbing is a useful criterion for specific differen- 
tiation. The anterior commissure is either rectimarginate, uniplicate or slightly 
sulcate. 

Internal characters. None of the specimens sectioned was particularly well preserved. 
Pedicle valve. Dental lamellae are present but are usually only weakly developed 
and invariably only attached, if at all, to the extreme posterior of the valve. The 
teeth are strong and usually inserted almost vertically into their sockets. 

Brachial valve. The median septum, if present, is low ; septalial plates are not 
developed. Inner and outer socket ridges are usually well differentiated. 

The crura are of the type distinguished by Wisniewska as arcuifer. She described 
them as having wide bases, being concave toward the middle and turning distally 
toward the ventral valve where they are terminated by " a sort of small crural 
plate in the shape of a hammer ". The author would agree with Wisniewska that 
the genus has a distinctive type of crura. The most distinctive feature about them 
as seen in transverse section would seem to be the way in which, from being initially 
flattened in the plane of articulation, by the development of vertical elements at 
their inner ends, they again become flattened but lying at right angles to the plane 
of articulation and slightly to the ventral side of it. 

Species. The following nominal species are attributed to the genus 
M. czenstochaviensis (Roemer) (1870, p. 247, pi. xxii, figs. 12-14) 
M. lineolata (Phillips) (as figured by Jacob and Fallot 1913, pp. 17-18, pi. i, 

figs. 9-14). 
M. rollieri Wisniewska (1931, pp. 59-60, pi. vi, figs. 10-11) 
M. striocincta (Quenstedt) (1852, p. 455, pi. xxxvi, fig. 24) 
M. strioplicata (Quenstedt) (1852, p. 455, pi. xxxvi, fig. 23) 
M. triloboides (Quenstedt) (1852, p. 455, pi. xxxvi, fig. 29) 



20 JURASSIC RHYNCHONELLIDS 

Distribution. The genus occurs most commonly in the Swabian and Franconian 
Jura ; outside those areas it has only definitely been recorded from Poland and the 
French Jura and in both these regions it is a very scarce form. Makridin (1964) 
does not record it from the Russian Platform, although it is stated to occur there in 
the Russian " Treatise ". 

Occurrence. The three species investigated in detail all have consistently well 
developed pedicle openings which, it is reasonable to assume, indicate a functional 
pedicle in the adult stage. This latter fact, coupled with the wide distribution of the 
species and the fact that they can occur in a variety of high energy environments, 
suggests that the mode of life could well have been one of attachment either to 
sponges or floating material. This hypothesis would account for the diversity of 
environment in which the genus occurs. 

Monticlarella czenstochaviensis (Roemer) 
(PI. i, fig. 4, text-fig. 8) 
1870 Rhynchonella Czenstochaviensis Roemer: 247-48, pi. 22, figs. 12-14. 
191 7 Rhynchonella Czenstochaviensis Roemer; Rollier: 116. 
1932 Monticlarella czenstochowiensis (Roemer); Wisniewska: 57-58, pi. 6, figs. 13-18. 

Emended diagnosis. Medium sized, subtriangular Monticlarella ; maximum 
width well toward anterior margin ; valves equally biconvex or with the pedicle 
valve the more inflated ; rectimarginate ; test covered by fine radial striae ; con- 
centric growth lamellae may be present ; crura arcuifer. 

Stratigraphical range. Lower Oxfordian — cor datum zone. 

Type specimen. Lectotype, here selected, the specimen figured by Roemer 
(1870) pi. xxii, figs. 12-14, from " Clarenberges " near Czenstochowa, Poland. 

Material. 5 specimens collected by Dr. D. A. B. Pearson from the Holy Cross 
Mts., Poland. 2 specimens obtained by Dr. D. V. Ager from the type locality. 

Dimensions of Figured Specimen. 

length thickness width 

1-02 cm 073 cm 1-13 cm 

n "O O "O n o o 




MONTICLARELLA CZENSTOCHOWIENSIS (Roemer) 
Oxlordian . Polond , J 1219/2 

Fig. 8. Transverse serial sections of Monticlarella czenstochaviensis (Roemer) kindly made 
available by Dr. D. V. Ager. Numerals represent distance in millimetres from pedicle 
umbo. 



FROM NORTHWESTERN EUROPE 21 

Distribution. The species has only been recorded from Poland. 

Remarks. The figured specimen and the series of serial sections were kindly 
made available by Dr. D. V. Ager. Lack of material precludes a full description but, 
from the limited material available and the specimens which have been figured, the 
species appears to be both consistent and distinctive. 

Although Wisniewska (1932) did not give any reason why she changed the name, 
it is presumed that she did so on grounds of incorrect transliteration. However, as 
this is not permissible under the International Code of Zoological Nomenclature, 
Article 32, (a) ii, the author has reverted to the original spelling of Roemer (1870). 



Monticlarella striocincta (Quenstedt) 
(PL 1, figs. 1-3, text-fig. 9) 

1852 Tevebratula striocincta Quenstedt: 455, pi. 26, fig. 24. 

1858 Tevebratula striocincta Quenstedt: 634, pi. 78, fig. 12. 

1871 Terebratula striocincta Quenstedt: 131, pi. 40, figs. 17-19. 

1885 Terebratula striocincta Quenstedt: 694, pi. 53, fig. 57. 

1886 Rhynchonella cf. striocincta (Quenstedt); Oppel and Waagen: 295. 
1917 Rhynchonella striocincta (Quenstedt); Rollier: 119. 

1932 Monticlarella striocincta (Quenstedt); Wisniewska: 58-59, pi. 6, fig. 12. 

Emended diagnosis. Small, subpentagonal Monticlarella ; anterior commissure 
rectimarginate or slightly sulcate ; pedicle valve much more inflated than the 
brachial ; ornament of fine striae passing anteriorly into coarser ribs ; crura 
arcuifer. 

Stratigraphical range. Oxfordian — Kimmeridgian : transversarium to muta- 
bilis zones. 

Type specimen. Lectotype, here designated, fig. 16, pi. 40 in Quenstedt (1871) 
from the " Malm gamma " of Lochen, Germany. 

Material. 90 specimens from the collection of the University of Tubingen ; 
2 specimens collected by the author from the southern French Jura ; 1 specimen 
from the collection of the Geology Department, Imperial College, London. 

Description. External characters. The shell is biconvex with the pedicle valve 
the more inflated ; this results largely from the distinctive manner in which the 
posterior part of the pedicle valve bends over so that the top of the very small, 
pointed beak lies almost at right angles to the lateral commissure. Although the 
beak lies in this position it is not sufficiently incurved to obscure the delthyrium. 
In the material studied, the small, disjunct deltidial plates mentioned by Wisniewska 
(1932) have not been observed and there is an open delthyrium. 

The test ornament is distinctive and consists posteriorly of numerous fine striae 
which pass anteriorly into well developed, rounded ribs formed by the fusion of 
several striae ; the striae continue in the troughs between the ribs as far as the 
anterior margin. Growth lamellae are usually present but tend to be sparsely 
distributed and only poorly developed. 



jurassic rhynchonellids 

Dimensions of Figured Specimens. 

length thickness 



o-8o cm 
092 cm 



o- 



cm 



0-55 cm 
067 cm 
o-6o cm 



width 
075 cm 
0-83 cm 
o-86 cm 



Internal characters. Pedicle valve. The dental lamellae are very much reduced 
and not attached to the ventral wall of the shell. The teeth are inserted at a 
relatively shallow angle, as seen in transverse section. 

Brachial valve. There is no median septum. The sockets are shallow and inner 
and outer socket ridges are scarcely differentiated. 




0-2 




0-3 





0-5 



Ky 



0-6 






10 



0-9 



0-7 





Fig. 9. 
alpha 
(X8). 



1-1 1-3 

Transverse serial sections of Monticlarella striocincta (Quenstedt) 



Lochengrundle, Germany. 



" Weiss Jura 
Specimen donated by the University of Tubingen. 



Distribution. The Swabian and Franconian Jura, Poland and the southern 
French Jura are the only areas from which the species has been definitely recorded. 
It is considered likely that M. striocincta could occur quite widely in the French Jura 
but has been overlooked on account of its small size. 



FROM NORTHWESTERN EUROPE 23 

Occurrence. Unfortunately Wisniewska (1932) does not give any details of the 
associated fauna or of the type of lithology from which her material was obtained 
and Quenstedt (1852, 1871) only says that it occurs with Terebratula lacunosa. 
However, the author visited the original Quenstedt locality of Lochengriindle near 
Tubingen and collected material from " Malm upper alpha " and " lower beta ". 
Both of these horizons are within the sponge reef facies and consist there of inter- 
bedded hard, argillaceous limestones and marls, occurring in units ranging from a 
few inches to two or three feet in thickness. M. striocincta, together with M. 
strioplicata, was found to occur predominantly in the marly units and was associated 
with a varied fauna of terebratulids, including the ribbed " pectunculus " types, 
echinoid plates and spines, small ammonites and sponges ; lacunosellid brachiopods 
were present but were concentrated in the harder limestone bands. 

In the southern French Jura (Ain), the species has been collected from the Bedded 
Virieu Limestone (Ager & Evamy, 1963) just north of Lac d'Armaille. The Bedded 
Virieu Limestone is here represented by well bedded calc-lutites with alternations of 
calcareous shale ; as well as M. striocincta, there is a rich fauna of lamellibranchs, 
terebratulids and the large asymmetrical rhynchonellid Torquirhynchia guebhardi. 
M. striocincta is one of the least abundant elements of the fauna, the majority of 
which suggests a high energy environment at or above wave-base (Evamy, 1963 
unpublished thesis). M. striocincta has also been collected in this area from the 
argillaceous Chavoley Beds and details of the occurrence are given under Lacunosella 
arolica. 



Monticlarella triloboides (Quenstedt) 
(PI. 1, figs. 5-7, text-fig. 10) 

1852 Terebratula triloboides Quenstedt: 455, pi. 26, fig. 29. 

1858 Terebratula triloboides Quenstedt: 643, pi. 78, fig. 13. 

1 87 1 Terebratula triloboides Quenstedt: 129, pi. 40, figs. 6-9. 

1876 Rhynchonella triloboides (Quenstedt); Loriol: 188, pi. 23, figs. 41-43. 

1885 Terebratula triloboides Quenstedt: 694, pi. 53, fig. 58. 

1904 Rhynchonella triloboides (Quenstedt); Loriol: 279, pi. 27, figs. 36-38. 

1913 Rhynchonella triloboides (Quenstedt); Jacob and Fallot: 34. 

1917 Rhynchonella triloboides (Quenstedt); Rollier: 119. 

1918 Stolmorhynchia triloboides (Quenstedt); Buckman: 46. 

1932 Monticlarella triloboides (Quenstedt); Wisniewska: 62-63, P 1 - 6, figs. 8-9. 

Emended diagnosis. Medium to large sized subpentagonal Monticlarella ; 
about 15-20 simple ribs ; no posterior smooth area ; radial striae only present in 
the troughs between the ribs ; uni plication in anterior commissure ; crura arcuifer. 

Stratigraphical range. Oxfordian — Kimmeridgian — Lower Volgian ; accu- 
rate records range from the transversarium to Gravesia zones, Quenstedt states that 
it is most abundant in the " Malm gamma ", i.e. cymodoce and mutabilis zones. 



24 JURASSIC RHYNCHONELLIDS 

Type specimen. Lectotype, here designated, fig. 8, pi. 40 of Quenstedt (1871) 
from the " Malm gamma " of Bosler, Germany. 

Material. 320 specimens from the collection of the University of Tubingen ; 
17 specimens collected by Dr. Enay from near Arinthod (Jura). 

Description. External characters. The shell is equally biconvex and has a 
small, pointed, erect beak. The oval pedicle foramen is flanked by disjunct deltidial 
plates which show a characteristic thickening around the actual opening. Beak 
ridges and interarea are both well developed. 

The ornament consists of about 20 rounded ribs, as counted at the anterior margin, 
which mostly originate at the umbones although a few also arise by dichotomy. 
The well marked uniplication in the anterior commissure usually carries about six of 
these ribs. A distinctive feature of the species is the radial striae which can be seen 
only in the troughs between the ribs of well preserved specimens ; these striae are 
usually most easily observable toward the anterior margin. The better preserved 
specimens also show a concentric ornament of very fine growth lines, which are 
usually more apparent posteriorly ; growth lamellae are rarely developed. 

The shell of many of the broken and worn specimens is seen to be of a very fibrous 
character ; whether this is its original condition or the result of post-burial recrys- 
tallisation, as evidenced by the loss of the fine detail, is not known. Although the 
shell fibres and the radial striae superficially resemble one another, the striae may be 
differentiated by the fact that they are more individually distinct and occur only in 
the troughs, whereas the fibres appear as a compact mass and may be visible on any 
part of the shell. If the top layer of the shell has been removed so that the fibrous 
condition is visible, this inevitably means that the striae will have been obliterated. 

Dimensions of Figured Specimens. 

length 
i -oo cm 
0-87 cm 
1-27 cm 
Internal characters. Pedicle valve. 
lamellae are relatively well developed, 
sockets, as seen in transverse section. 

Brachial valve. There is no m 
are as in generic description. 

Distribution. The species is known from the Swabian and Franconian Jura, 
Poland and the central French Jura. The only other record is that of Jacob and 
Fallot (1913) from the " Portlandien superieur " of Chomerac (Ardeche). 

Occurrence. The author has not collected the species and it is only briefly 
recorded in the literature. Quenstedt (1871) noted that it occurred with Terebratula 
lacunosa. The horizon cited by Jacob and Fallot appears to yield a very varied 
rhynchonellid fauna, some of which is almost certainly referable to the genus 
Lacunosella Wisniewska. The material of Dr. Enay came from a fine grained marl 
from which specimens of Acanthorhynchia (Echinirhynchia) lorioli were also obtained. 



thickness 






width 


o-68 cm 






1 -oo cm 


0-59 cm 






0-87 cm 


0-84 cm 






1-26 cm 


The shell : 


is 


quite 


thick and the dental 


i. The teeth 


are inserted vertically into the 


in septum or 


ri< 


ige de' 


^eloped ; other details 



FROM NORTHWESTERN EUROPE 




2-5 2-3 

Fig. io. Transverse serial sections of Monticlarella triloboides (Quenstedt). " Malm 
gamma", Germany. Specimen donated by the University of Tubingen. (x8). 



26 JURASSIC RHYNCHONELLIDS 

Monticlarella strioplicata (Quenstedt) 
(PI. i, figs. 8-10, text-fig. n) 

1852 Terebratula strioplicata Quenstedt: 455, pi. 26, fig. 23. 

1858 Terebratula strioplicata Quenstedt: 635, pi. 78, fig. 24. 

1871 Terebratula strioplicata Quenstedt: 132, pi. 40, figs. 24-32. 

1885 Terebratula strioplicata Quenstedt: 694, pi. 53, fig. 59. 

1909 Rhynchonella strioplicata (Quenstedt); Simionescu: 40, 41, 94, pi. 7, fig. 18. 

1913 Rhynchonella strioplicata (Quenstedt); Jacob and Fallot: 17. 

1917 Rhynchonella strioplicata (Quenstedt); Rollier: 123. 

1917 Rhynchonella strioplanala (Quenstedt); Rollier: 123. 

1917 Rhynchonella pauciplicata Rollier: 123. 

1917 Rhynchonella tenuiplicata Rollier: 123. 

1917 Rhynchonella furcatella Rollier: 123. 

1932 Monticlarella strioplicata (Quenstedt); Wisniewska: 60-62, pi. 6, figs. 1-7, text-fig. 18. 

Emended diagnosis. Medium sized, subpentagonal to slightly elongate Monti- 
clarella ; rectimarginate or with a low uniplication ; ornament of fine striae poster- 
iorly passing anteriorly into short coarse ribs ; crura arcuifer. 

Stratigraphical range. Oxfordian — Kimmeridgian — Lower Volgian ; trans- 
versarium to Gravesia zones. 

Type specimen. Lectotype, here designated, fig. 25, pi. 40 of Quenstedt (1871) 
from the "Malm gamma " of Lochen, Germany. 

Material, ii specimens from the collection of the University of Tubingen ; 
22 collected by the author from Swabia and Franconia. 

Description. External characters. The shell is biconvex with the pedicle valve 
slightly more inflated than the brachial. The beak is small, sharply pointed and 
erect. The oval, hypothyridid pedicle opening is small and is flanked by two small, 
discrete deltidial plates. Beak ridges and interarea are only poorly developed. 

Posteriorly the ornament consists of very fine striae, while on the anterior half or 
third of the shell are developed five or six relatively coarse ribs. The fine striae 
continue to the anterior of the shell only in the troughs between the ribs. The 
outline varies from subpentagonal to suboval, specimens having the former shape 
being wider and having more ribs ; the length is usually greater than the width. 
There appears to be a complete gradation between the two forms at all stratigraphical 
levels for which there is material available ; however, the material is relatively 
limited and it is not possible to say whether the proportions of the two forms change 
or have any stratigraphical value. 

Dimensions of Figured Specimens. 

length thickness width 

1-18 cm o-8o cm 1-13 cm 

1-04 cm o-68 cm 098 cm 

105 cm 070 cm 078 cm 

Internal characters. Pedicle valve. Dental lamellae are present but only poorly 
developed and attached to the shell wall only at their posterior extremities. The 
teeth are strong. 



FROM NORTHWESTERN EUROPE 



2 7 





0-5 



O 



0-7 




0-9 







1-7 



1-5 



1-3 




2-2 



Fig. ii. Transverse serial sections of Monticlarella strioplicata (Quenstedt) . " Malm 
gamma". Barrenberg, Germany. (x8). 



28 J IRASSIC RHYNCHONELLIDS 

Brachial valve. A long, well developed median ridge is present. The hinge plates 
are flat. Both the inner and outer socket ridges are clearly differentiated. 

Distribution. Apart from the Swabian and Franconian Jura and Poland, the 
only other records of the species are those of Jacob and Fallot (1913) from Mont 
Crussol and from Chomerac (Ardechc). A possible specimen of M . strioplicata was 
figured by Struckman (1878) from near Hanover. 

Occurrence. The species has been collected by the author from only two locali- 
ties, namely, at Lochengrundle in the Swabian Jura and at Barrenberg in Franconia. 
The former locality has been described in the remarks on the occurrence of M. 
striocincta. The latter locality is of " Malm gamma " age and is again in the sponge 
reef facies. The fauna included many lacunosellids, terebratulids, small sponges, 
ammonites and occasional belemnites and echinoids, with M. strioplicata one of the 
least abundant faunal elements. This locality is further described under Lacuno- 
sella sparsicosta. 

Family WELLERELLIDAE Likarev in Rzhonsnitskaya, 1956 

Subfamily LACUNOSELLINAE Smirnova, 1963 

Genus LACUNOSELLA Wisniewska, 1932 

1932 Lacunosella Wisniewska: 30-32. 

i960 Lacunosella Wisniewska; Makridin: 254. 

1965 Lacunosella Wisniewska; Ager: 608. 

Type species. Rhynchonella arolica Oppel, by original designation. 

Emended diagnosis. Medium to large rhynchonellids ; outline usually sub- 
pentagonal ; ribs bifurcate ; beak small ; beak ridges and interarea poorly 
developed ; shell massive ; very reduced dental lamellae ; crura falcifer. 

Stratigraphical range. Oxfordian — Kimmeridgian — Lower Volgian ; upper 
cordatum to Gravesia spp. zones. 

Description. External characters. The shell is either equally biconvex or with 
the brachial valve the more inflated. The test frequently has a characteristic and 
distinctive translucent, pearly appearance. Although the outline is usually sub- 
pentagonal, it may be strongly trilobate as in L. trilobata. The genus is usually 
symmetrically developed with a uniplication in the anterior commissure, however, 
this feature is frequently displaced by asymmetrical development and completely 
asymmetrical, bilobed variants are produced by L. trilobata. 

The ribbing is variable and in different species the rib count at the anterior com- 
missure varies between five and twenty-five. One of the most distinctive external 
features is that while not all the ribs bifurcate, this character is invariably shown by 
some of them. 

A small suberect beak is present, as are beak ridges and an interarea although both 
the latter are only poorly developed. The submesothyridid pedicle opening is 
flanked by conjunct deltidial plates. Concentric ornament, which is neither strongly 
developed nor consistently present, comprises fine growth lines and weakly developed 
growth lamellae. 



FROM NORTHWESTERN EUROPE 29 

Internal characters. Pedicle valve. The most striking features are the massive 
shell and the very reduced dental lamellae ; lateral cavities are hardly developed. 
Pedicle collars were not present in all the species sectioned but as it seems probable 
that these are only developed in the adult specimens, their absence may be the result 
of sectioning slightly immature material. The teeth are strong and frequently show 
some degree of crenulation ; lateral denticulae are usually present. In fully adult 
specimens the muscle attachment areas can be clearly seen as quite deep depressions 
in the shell. 

Brachial valve. There is no septalium and the median septum, if present, is very 
much reduced. The hinge plates are divided and, almost as soon as the cavity of 
the brachial valve is exposed during serial sectioning, the long falcifer crura can 
be seen hanging down from them into the brachial valve. The presence of falcifer 
crura distinguishes Lacunosella from all other Upper Jurassic genera. Inner and 
outer socket ridges are both well-developed. 

Species. The following nominal species are attributed to the genus : 

L. amstettensis (Fraas) (1858, pp. 107-8) 

L. arolica (Oppel) (1865, p. 312) 

L. blanowicensis Wisniewska (1932, pp. 40-1, pi. iii, figs. 1-5) 

L. cracoviensis (Quenstedt) (1871, p. 136, pi. xl, fig. 43) 

L. dilatata (Rollier) (1917, p. 134) 

L. kozlowskii Wisniewska, (1932, pp. 54-55, pi. v, figs. 1-7) 

L. monsalvensis (Gillieron) (1873, p. 244, pi. x, figs, n-12) 

L. multiplicata (Zieten) (1832, p. 55, pi. xli, fig. 5) 

L. polita (Quenstedt) (1871, p. 126, pi. xxxix, fig. 106) 

L. prosimilis (Rollier) (1917, p. 174) 

L. pseudodecorata (Rollier) (1917, p. 139) 

L. selliformis (Lewinski) (1908, p. 437, pi. xxii, fig. 8-1 1) 

L. silicea (Quenstedt) (1871, p. 129, pi. xxxix, fig. 107) 

L. siemiradzkii (Wisniewska) (1932, pp. 48-49, pi. iii, figs. 21, 22) 

L. sparsicosta (Quenstedt) (1858, pp. 633-34, P^ lxxviii, figs. 12-22) 

L. subsimilis (Schlotheim) (1820, p. 264) 

L. trilobata (Zieten) (1832, p. 56, pi. xlii, fig. 3) 

L. trilobataeformis Wisniewska (1932, pp. 46-48, pi. iv, figs. 1-11) 

L. vaga sp. nov. 

L. visulica (Oppel) (1866, p. 295). 

Distribution. The genus has been recorded throughout the French, German 
and Swiss Jura, from Poland and from Stramberk, Czechoslovakia. 

Occurrence. With the exception of the occurrence of L. vaga on the Col du 
Chat (Savoie) the genus is always associated with sponges. This association has 
been commented on by Quenstedt (1871), Ager and Evamy (1963) and Ager (1965a). 
The genus is most abundant and varied in the Swabian and Franconian Jura where 
it occurs in the bedded limestones associated with the sponge reefs. 

The massive shell and consistent presence of a pedicle opening, presumably 
indicating a functional pedicle, seem to be adaptations to a high energy environment 



30 JURASSIC RHYNCHONELLIDS 

around the reefs. The persistent occurrence of sponges with Lacunosella is so 
striking that it suggests that the genus was in some way dependent on them. 
Several times while collecting material, the author found specimens of Lacunosella 
actually in contact with sponges. While this may have been coincidence, it could 
also have been the result of the rhynchonellids actually having been attached to the 
sponges. It is difficult to see what particular benefit this would have been and why 
Lacunosella should have restricted itself to attachment on that particular ' host '. 
At the locality mentioned above where sponges are not present, L. vaga was obtained 
from bedded limestones associated with " reef " limestones. 



Lacunosella arolica (Oppel) 
(PI. i, figs. 11-12, text-figs. 12-13) 

1864 Rhynchonella lacunosa (Schlotheim) ; Heer: 137, fig. 87. 

1865 Rhynchonella Arolica Oppel: 312. 

1866 Rhynchonella Arolica Oppel; Oppel and Waagen: 294. 

1867 Rhynchonella Arolica Oppel; Moesch: 310-31 1, pi. 7, fig. 9. 
1871 Terebratula lacunosa Arolica Quenstedt; 127, pi. 39, figs. 95-96. 
1885 Terebratula lacunosa Arolica Quenstedt: 693, pi. 53, fig. 60. 

1890 Rhynchonella lacunosa (Quenstedt) var. Arolica Oppel; Haas: 48-50, pi. 3, figs. 1-22, pi. 
4, figs. 1-5. 

191 7 Rhynchonella Arolica Oppel; Rollier: 134. 

1918 Stolmorhynchia arolica (Oppel); Buckman: 46. 

1932 Lacunosella arolica (Oppel); Wisniewska : 33-38, pi. 2, figs. 1-18. 

Emended diagnosis. Medium to large sized, relatively flattened Lacunosella ; 
outline subpentagonal ; 9-1 1 low, subangular ribs; crura falcifer. 

Stratigraphical range. Oxfordian — upper cordatum and transversarium zones. 

Type specimen. In his original description of the species Oppel cited the figure 
in Heer (1864) listed above. 

Material. 16 specimens collected by the author from the southern French 
Jura in the British Museum (Nat. Hist.) nos. BB. 45630-45639, BB. 45678-45681 ; 
43 specimens collected by Dr. Enay from the central French Jura ; 12 specimens 
from the " Birmensdorfer-Schichten " of Switzerland donated by the Natur- 
historisches Museum, Basel. 

Description. External characters. The shell is biconvex with the brachial 
valve slightly the more inflated ; however, both valves are relatively flattened com- 
pared with other species of the genus. The beak is small and suberect ; beak 
ridges and interarea are hardly differentiated. The small, submesothyridid pedicle 
opening is flanked by conjunct deltidial plates. 

The radial ornament consists of about eleven low, subangular ribs, the majority of 
which originate at the umbones although bifurcation is not uncommon. Numerous 
fine growth lamellae are characteristically present. The anterior commissure has a 
well-marked uniplication which usually carries three or four ribs ; however, it is 
often distorted to some degree by asymmetrical development. 



from northwestern europe 31 

Dimensions of Figured Specimens. 

length thickness width 

3-30 cm 235 cm 378 cm 

2-86 cm 2-20 cm 3-22 cm 

Internal characters. The internal characters have been investigated using material 
collected near Virieu-le-Grand (Ain) in the southern French Jura. Although not 
topotypic, the material is of the correct age and is not far from Trept (Isere), a 
locality mentioned by Oppel (1866) in his original description of the species. 

Pedicle valve. A well developed pedicle collar is present in adult specimens. The 
shell is massive and while lateral cavities are present they are of a very reduced 
nature and bounded by slender dental lamellae. Although the dental lamellae are 
very slight, the teeth are strongly supported by the thick shell wall. 

Brachial valve. There is a low median septum ; other characters as in generic 
description. 

Distribution. The species appears to occur quite abundantly throughout the 
French and Swiss Jura and in Poland. That L. arolica has not been recorded from 
either the Swabian or Franconian Jura is attributed to the fact that the " Sponge 
reef " fades did not develop in that area until the lower Kimmeridgian. 

Occurrence. The material collected by the author came from the Chavoley 
Beds (Ager & Evamy, 1963) near Virieu-le-Grand (Ain). These beds, consisting of 
calcareous marls and shales, represent the only argillaceous interruption in the 
succession of carbonate sediments which constitute the Upper Jurassic of the 
southern French Jura. The associated fauna included hexactinellid sponges and 
rare specimens of Monticlarella striocincta. L. arolica was the commonest element 
in the fauna and is also abundant in the "Birmensdorfer Schichten " of the type area 
in Switzerland. 

Lacunosella vaga sp. n. 
(PI. 2, figs. 1-3, text-fig. 14) 

1932 ?Septaliphoria pinguis (Roemer) ; Wisniewska: pi. 1, figs. 13, 14, NON figs. 1-11, 15-20. 

Name. Latin vagus-a-um : wandering, the species has a wide geographical 
distribution. 

Diagnosis. Medium-sized, symmetrical, sub-pentagonal Lacunosella ; beak 
small ; 16-20 subangular ribs ; no posterior smooth area ; crura falcifer. 

Stratigraphical range. Kimmeridgian — only accurately recorded from the 
upper mutabilis and lower pseudomutabilis zones. 

Type specimen. Holotype, no. BB. 44135, obtained from mutabilis zone at 
Sermerieu (Isere), France. 

Dimensions of holotype. Length 2-69 cm, thickness 200 cm, width 3-02 cm. 

Material. 43 specimens from type locality ; 22 specimens Col du Chat (Savoie) ; 
28 specimens from Franconia. All except the figured specimens are in the author's 
collection, British Museum (Nat. Hist.) nos. BB. 45668-73, BB. 45756-61. 



32 



JURASSIC RHYNCHONELLIDS 




6-3 



5 6 



Fig. 12. Transverse serial sections of Lacunosella arolica (Oppel). Oxfordian. Chavoley 
Beds. Virieu-le-Grand (Ain), France. (X3). 



FROM NORTHWESTERN EUROPE 



33 



Description. External characters. L. vaga is weakly trilobate with a low, 
symmetrical uniplication in the anterior commissure. The brachial is the more 
inflated valve and the species is generally characterised by a rather globose appear- 
ance. Concentric ornament is only rarely present and consists of weak growth 
lamellae and numerous fine growth lines. The beak is typically small and sub- 
erect ; conjunct deltidial plates delimit the small submesothyridid pedicle opening. 
Beak ridges and an interarea are present but only poorly developed. 




2 5 




4 4 




3 9 




3 5 




5 




5 7 



Fig. 13. Transverse serial sections of Lacunosella arolica (Oppel). Oxfordian. Chavoley 
Beds. Virieu-le-Grand (Ain), France. (X3). 

Dimensions of Figured Specimens. 



length 

2-69 cm 
2-84 cm 
278 cm 



thickness 

2-oo cm 
1-84 cm 
i-go cm 



width 

3-02 cm 
3-06 cm 
2-97 cm 



34 JURASSIC RHYNCHONELLIDS 

Internal characters. Pedicle valve. Shell massive and the muscle impressions 
deeply inserted. The teeth are strongly developed and lateral denticulae are present. 

Brachial valve. Typical lacunosellid falcifer crura are developed. Very strong 
inner and outer socket ridges are present. 

Distribution. The species has been collected by the author from the southern 
French Jura in the Departments of Savoie and Isere and from the Franconian Jura. 
Two of the specimens from Poland figured by Wisniewska (v. synonymy) as Septali- 
phoria pinguis are externally identical with L. vaga and, as they also come from the 
same horizon, it seems very likely that Wisniewska misattributed them. This 
suggestion is borne out by the fact that Royzicki (1948) recorded 5. pinguis sensu 
Wisniewska from sponge reefs. If this is the case, then the geographical distribu- 
tion of the species would be considerably extended. 

Occurrence. At the type locality, L. vaga is associated with Torquirhynchia cf. 
astieriformis and details of the locality are given under that species. The specimens 
collected from the Col du Chat (Savoie) were from fine-grained, bioclastic limestones. 
This record is interesting in that it is the only occurrence of Lacunosella known to 
the author where the genus is not associated with sponges ; many of the specimens 
were crushed and the only other fossils collected from the locality were terebratulids. 
The Franconian material was found with the typical association of sponges, ammo- 
nites, terebratulids and lamellibranchs at the base of a small reef. 



Lacunosella cracoviensis (Quenstedt) 
(PI. 2, figs. 5-6, text-fig. 15) 

1871 Terebratula trilobata cracoviensis Quenstedt: 136, pi. 40, fig. 43. 

1871 Terebratula lacunosa subsimilis (Schlotheim) ; Quenstedt: 129, pi. 39, fig. 78. 

1917 ? Rhynchonella cracoviensis (Uhlig); Rollier: 172. 

1917 Rhynchonella prosimilis Rollier: 174. 

1932 Lacunosella cracoviensis (Quenstedt); Wisniewska: 41-46, pi. 2, figs. 6-20. 

Emended diagnosis. Medium-sized, subpentagonal Lacunosella ; prominent 
narrow, suberect beak ; 17-25 fine, bifurcating ribs ; crura falcifer. 

Stratigraphical range. Upper Oxfordian — Lower Kimmeridgian ; in Swabia 
and Franconia it has been recorded from the " beta " and " gamma Malm " while 
Wisniewska gives its range in Poland as " Sequanien ". 

Type specimen. Lectotype, here selected, fig. 43, pi. 60 in Quenstedt (1871), 
from the " Coralrag " of Przegorzaty, Poland. 

Material. 7 specimens, collection of Tubingen University from Swabia ; 21 
specimens, author's collection from Franconia and Swabia, in the British Museum 
(Nat. Hist.) nos. BB. 45650-67, BB. 45674-77. 

Description. External characters. L. cracoviensis is a medium-sized Lacuno- 
sella ; although the outline is subpentagonal, it shows considerable variation and 
either length or width may be the greater dimension. 

The subangular ribs are relatively fine, commence at the umbones and are often 
bifurcate. The anterior commissure is usually symmetrical with a low uniplication 



FROM NORTHWESTERN EUROPE 



35 



developed but not uncommonly it shows varying degrees of asymmetry. There are 
usually 8-9 ribs on the fold. 




55 




6 9 




6 3 



Fig. 14. Transverse serial sections of Lacunosella vaga sp. nov. Topotype. Kimmeridgian- 
mutabilis zone. Sermerieu (Isere), France. (X3). 



3" 



JURASSIC RHYNCHONELLIDS 



The suberect beak is narrow and characteristically pronounced. The interarea 
and beak ridges are scarcely developed. The submesothyridid pedicle opening is 
small and flanked by conjunct deltidial plates. 

Dimensions of Figured Specimens. 

length thickness width 

2-62 cm 1-40 cm 2-50 cm 

276 cm 1-97 cm 2 94 cm 

Internal characters. Pedicle valve. The shell is massive ; lateral cavities are not 
developed as the dental lamellae do not appear to touch the ventral shell wall. A 
pedicle collar is present. The teeth are strongly developed and lateral denticulae 
are present. 

Brachial valve. The falcifer crura are exceptionally wide distally. Inner and 
outer socket ridges are very strongly developed. 




47 



Fig. 15. Transverse serial sections of Lacunosella cracoviensis (Quenstedt). 
gamma". Wiirgau, Germany. (X4). 



" Malm beta/ 



FROM NORTHWESTERN EUROPE 37 

Distribution. The species is recorded from Poland, and the Swabian and 
Franconian Jura. 

Occurrence. Occurs in association with sponge reefs, together with the usual 
fauna of ammonites and terebratulids. 

Remarks. The name cracoviensis was first used by Quenstedt (1871) in describing 
material from the area of Krakow, Poland ; the specimen he then figured, pi. 60, 
fig. 43, has been selected as lectotype. The species has been well figured by 
Wisniewska (1932) and the author would agree with her in placing in synonymy the 
specimen figured by Quenstedt (1871), pi. 39, fig. 78, as Terebratula lacunosa sub- 
similis from the " Malm beta ". This latter specimen undoubtedly belongs to this 
rather variable species and other German material from this horizon can be matched 
exactly with the specimens figured by Wisniewska from Poland. 

The name subsimilis was coined by Schlotheim (1820) for material from the 
" Hornstein ", near Amberg, which is generally assumed to be " Malm epsilon " 
in age. Quenstedt figured a specimen of subsimilis of this age (pi. 39, fig. 108). 
Rollier (1917) proposed the name prosimilis for the specimen figured by Quenstedt 
from the " Malm beta " in order to differentiate it from subsimilis of Schlotheim. 
No specimens from the " Malm epsilon " referrable to the Schlotheim species have 
been examined. It has been decided, in view of the stratigraphic diastem, namely 
" Malm upper gamma " and " delta " which separates it from L. cracoviensis, 
together with the fact that lack of material makes taxonomic revision impossible, 
to continue to regard subsimilis as a separate species restricted to the " Malm 
epsilon ". 

Lacunosella sparsicosta (Quenstedt) 
(PI. 2, figs. 7-8, text-fig. 16) 

1858 Terebratula lacunosa sparsicosta Quenstedt: 633-34, pi. 78, figs. 19-22. 

1871 Terebratula lacunosa sparsicosta Quenstedt: 126-28, pi. 39, figs. 92-93, ?94- 

1890 Rhynchonella lacunosa var. sparsicosta (Quenstedt); Haas: 51-52, pi. 6, figs. 9-12, 15-17. 

1917 Rhynchonella sparsicosta (Quenstedt); Rollier: 140. 

Emended diagnosis. Small to medium-sized, subpentagonal Lacunosella ; 6-7 
ribs anteriorly of which 2-3 lie on the fold ; posterior smooth area present ; crura 
falcifer. 

Stratigraphical range. It appears to be restricted to the " Malm gamma ", 
which is equivalent to the cymodoce zone of the Kimmeridgian. 

Type specimen. Lectotype, here selected, fig. 93, pi. 39 in Quenstedt (1871), 
from the " Malm gamma " of Kasbuhl, Germany. 

Material. 10 specimens, author's collection from Franconia in British Museum 
(Nat. Hist.) nos. BB. 45640-49. 35 specimens, collection of Tubingen University 
from Swabia ; 13 specimens, collection of the Naturhistorisches Museum, Basel, 
from the Schaffhausen area. 

Description. External characters. L. sparsicosta is one of the smallest species 
in the genus ; it is subpentagonal in outline and has a strong uniplication in the 
anterior commissure. It is biconvex with the brachial valve considerably more 



38 JURASSIC RHYNCHONELLIDS 

inflated than the pedicle. The most frequent number of ribs developed is seven, of 
which three lie in the fold ; these latter ribs are coarse and subangular, contrasting 
with the lower more rounded ones occurring laterally. There is a posterior smooth 
area, as the ribs do not originate from the umbones ; the central ribs are, however, 
developed much further posteriorly than the lateral ones. The ribs are mostly 
simple. Growth lamellae are usually present but are very variably developed. 

The beak is small and suberect. A small, submesothyridid pedicle opening is 
flanked by small, conjunct deltidial plates. The interarea and beak ridges are only 
very poorly differentiated. 

Dimensions of Figured Specimens. 

length thickness width 

i-6i cm i 03 cm 1-73 cm 

1-98 cm 193 cm 2-23 cm 

Internal characters. Pedicle valve. There is a well-developed pedicle collar. 
The shell is massive and the dental lamellae are only poorly developed so that lateral 
cavities are hardly present. The delthyrial cavity is subquadrate and ventrally the 
depressions of the muscle attachment areas can be clearly seen. The teeth are 
strong and lateral denticulae are present. 

Brachial valve. Inner and outer socket ridges are both well-developed. Other 
features are as described for the genus. 

Distribution. The species ranges from the northern Franconian Jura through 
the Swabian Jura to the Schaffhausen area of Switzerland. 

Occurrence. The author has only collected the species from one locality, 
namely Barrenberg in Franconia. The exposure there was not very good but 
appeared to represent perireefal limestones which had been broken up and the 
resulting angular blocks recemented. L. sparsicosta was the commonest element in 
a rich fauna which also included sponges, ammonites, L. cracoviensis, Monticlarella 
strioplicata ' normal ' terebratulids and the sulcate terebratulid Nucleolata. 

Lacunosella trilobata (Zieten) 
(PI. 3, figs. 3-6) 

1832 Terebratula trilobata Munster; Zieten: 56, pi. 42, fig. 3. 

1832 Terebratula inaequilatera Goldfuss; Zieten: 56, pi. 42, fig. 4. 

1845 Terebratula trilobata Zyszner: pi. 5, figs. 1-5. 

1870 Rhynchonella trilobata Moesch: 145, pi. 14, fig. 33. 

1871 Terebratula trilobata Quenstedt: 134-36, pi. 40, figs. 35-39, ?figs. 40-44. 
1890 Rhynchonella trilobata (Zieten); Haas: 55-58, pi. 7, figs. 3-6. 

1917 Rhynchonella trilobata (Zieten); Rollier: 175. 

1917 Rhynchonella inaequilatera (Zieten); Rollier: 176. 

1918 Stolmorhynchia trilobata (Zieten); Buckman: 46. 
1932 Lacunosella trilobata (Zieten); Wisniewska : 32. 

Emended diagnosis. Medium-sized, elongated Lacunosella ; very strongly 
trilobate or asymmetrical ; 17-24 rather shallow, subangular ribs of which 6-9 lie 
on the fold. 



FROM NORTHWESTERN EUROPE 




5 5 



Fig. i 6. Transverse serial sections of Lacunosella sparsicosta (Quenstedt). 
ma". Barrenberg, Germany. (X4). 



" Malm gam- 



Stratigraphical range. The species has only been recorded from the " Malm 
epsilon " i.e. Lower Volgian. 

Type specimen. Lectotype, here selected, the specimen figured by Zieten, pi. 42, 
fig. 3. Although Zieten attributes the species to Miinster, the reason for this is not 
known as there does not appear to be any publication of Munster's prior to Zieten in 
which the species is mentioned. Sherborn (1930) throws no light on the matter, 
attributing speciosa to " von Miinster in Zieten ". Similarly, there seems no reason 
for his attribution of Terebratula inaequilatera to Goldfuss. 

Material. 2 specimens from the collection of the University of Tubingen ; 1 
specimen from the collection of the Naturhistorisches Museum, Basel, coming from 



4 o JURASSIC RHYNCHONELLIDS 

the " Kimmeridgien of Wiirttemberg "; 5 specimens in author's collection from the 
" Malm epsilon " of Franconia, in the British Museum (Nat. Hist.) nos. BB. 45750-54. 

Description. External characters. The most striking feature of the species is 
the very marked trilobation shown by most specimens. The species appears to be 
unique, at least in the upper Jurassic, in that it shows a complete gradation of forms 
between trilobate symmetrical and bilobed strongly asymmetrical forms. Especially 
in the trilobate specimens, the brachial valve tends to be considerably the more 
inflated. 

The bifurcating ribs vary considerably in number and development ; generally 
their coarseness is inversely proportional to the number present. The concentric 
ornament, when present, consists of numerous fine growth lines. The beak is small 
and flanked by weakly developed beak-ridges ; the interarea is slightly incurved. 
The small pedicle opening is limited by conjunct deltidial plates. 

Dimensions of Figured Specimens. 

length thickness width 

3-06 cm 2-25 cm 2-80 cm 

2-85 cm 2-05 cm 3-19 cm 

2-85 cm 1-63 cm 278 cm 

3-46 cm 2-55 cm 3-21 cm 

Internal characters. These were not investigated as the author's collection 
contains only a very limited number of sufficiently well preserved specimens. 

Distribution. The species appears to be restricted to the Swabian and Fran- 
conian Jura and the Schaffhausen area of Switzerland. 

Occurrence. The specimens collected by the author were obtained from well- 
bedded, fine-grained limestones ; they were the most abundant element in a fauna 
which also included ammonites and sponges. The fact that so few complete speci- 
mens were obtained is a result of the extremely hard, brittle nature of the host rock. 

Lacunosella sp. 

Specimens have been selected from material in the collection of the University of 
Tubingen in order to conform as nearly as possible to the specimens figured by 
Quenstedt (1858, 1871) as Terebratula lacunosa multiplicata (pi. 3, fig. 1) and T. 
lacunosa decor ata (= Rhynchonella pseudodecorata of Rollier) (pi. 1, fig. 13). The 
author has not personally collected material referable to this species and as there is 
only a very limited amount available at Tubingen, it has not been possible to give 
full descriptions. " Rhynchonella " pseudodecorata is of particular interest as it has 
been selected by Makridin (1955) as the type of his genus Isjuminelina (=Isjumin- 
ella). As this species seems to be externally a typical member of Lacunosella, it is 
considered that Makridin probably misidentified his material as his description of 
its internal structures excludes it from that genus. 

Also figured is a specimen of L. selliformis (Lewinski) (pi. 2, fig. 4) from the 
Oxfordian of Poland and a specimen of L. trilobataeformis Wisniewska (pi. 3, fig. 2). 
The latter is strongly homoeomorphic with Somalirhynchia moeschi and these two 
species appear to have been frequently confused in the past. 



FROM NORTHWESTERN EUROPE 41 

Family RHYNCHONELLIDAE Gray, 1848 

Subfamily RHYNCHONELLINAE Gray, 1848 

Genus RHYNCHONELLA Fischer, 1809 

1809 Rhynchonella Fischer: 35. 

1918 Rhynchonella Fischer: Buckman, 57. 

1920 Eurhynchonella Leidhold: 352. 

1932 Rhynchonella Fischer; Wisniewska: 13. 

1957 Rhynchonella Fischer; Ager: 1-15. 

i960 Rhynchonella Fischer, Makridin: 252. 

1964 Rhynchonella Fischer; Makridin: 111-13. 

19656 Rhynchonella Fischer; Ager: H610. 

Remarks. It is not proposed to revise the generic description as the type species, 
R. loxiae, does not occur in western Europe and within this area the genus is only 
represented by rare and very variable " species ". R. loxiae has recently been 
described and figured by Ager (1957) and Makridin (1964). 

In Britain, there are only two described Jurassic " species " and very little 
material which can be referred to Rhynchonella s.s. The two species are R. port- 
landica Blake and R. subvariabilis Davidson ; the type of the former was collected 
from the Lower Portland Sand of " Black Ven " Portland, and of the latter from the 
Kimmeridge Clay of Potterne, Wilts. Most of the material referable to Rhynchonella 
has been obtained from the " Rhynchonella Marls " and various horizons in the 
Portland Sand exposed on the Dorset coast. Apart from this, there are two speci- 
mens from the Lydite Bed of Long Crendon, Buckinghamshire, and six specimens 
from the Kimmeridgian Boulder Beds of Sutherland. Material from these horizons 
is very scarce and the author is indebted to the curators of the Geological Museum, 
Oxford, the Sedgewick Museum, Cambridge, and the B.M.(N.H.), for allowing him 
to use the specimens in their collections. 

An interesting series of specimens collected by Prof. M. R. House from the Upper 
Blackmore Beds, Portland, is figured on plate 4. As these were collected from the 
same horizon and locality, it can be reasonably assumed that they represent a single 
variable species. The variation ranges, however, between end-members, which, if 
they occurred separately and without intermediate forms, would undoubtedly be 
referred to R. portlandica and R. subvariabilis. This suggests that perhaps all the 
British material from this area should be referred to a single species. However, it is 
considered that further stratigraphically accurately located material is needed before 
the full range and variation can be determined ; until then it seems pointless to try 
and define taxonomic groups within this plexus of forms which is at present repre- 
sented by less than thirty specimens, of which few are complete. It is interesting 
to note that Blake (1880) in proposing his species R. portlandica, commented, 
" One can scarcely feel quite certain whether this is a distinct species from R. 
subvariabilis ". 

Specimens were collected by the author from the Boulder Beds of Sutherland, 
and appear to be the first record of the genus from that area. Again, as with the 
material from the Dorset coast, it does not seem helpful to propose further taxa 
until more specimens are available. 



4-2 



J URASSIC RHYNCHONELLIDS 



Makridin (1964) figured several species and subspecies of Rhynchonella and a table 
giving the stratigraphical ranges of these is given below (text-fig. 17) ; the strati- 



3> 

H 

I 
</l 

n 
n 

n 
</> 



1 

Q 
Q. 

a 


1} 

c 

a 

1 

O 



» 


(n 
O 

"■< 
Q 

Q 
c 
% 




*( 

Q 




* 

O 
Q 


3D 




3 


33 

O 

C 

« 
« 
O 

Q 


3) 


c 

* 

c 


O 

I* 

*- 


O 
C 

% 
O 

c 

% 


JO 

c 
</> 
in 

> 

z 

w 

TJ 

m 


m 


pectinolus 
r otundo 




s 


















pseudoscy thy co 


pallosioides 


s 


\ 


















ponder i 


olbani 


\ 
\ 
\ 
\ 


\ 


















virga tus 


gorei 


\ 


\ 

\ 


















rosonovi 


gig ant eus 


\ 

\ 

\ 




















nikitini 


















\ 






f ulgens 








\ 
















su bdi tus 
























nodiger 
























rjasanensis 












\ 












sten omphala 












1 1 











Fig. 17. The stratigraphical ranges of the Russian and British species of Rhynchonella s.s. 
Details of the Russian species are taken from Makridin (1964). 



FROM NORTHWESTERN EUROPE 43 

graphical range of the British Rhynchonellas s.s. is also indicated. Some of the 
forms figured by Makridin from the Russian Platform seem to be quite close to 
British specimens although the variation of the genus was such that none of the 
forms from these two areas is considered to be conspecific. 

The most distinctive British element seems to be the form with a long, posterior 
smooth stage and strong, short ribs anteriorly. This form is represented in the 
specimens collected by Prof. M. R. House, mentioned above, and is the basis on 
which such species as R. rivelensis and R. triplicosa are included in the genus. These 
species are all characterized by the above features and seem to comprise a persistent 
stock which is first seen in the basal Callovian and continues to the Volgian. It 
appears that the form represented by loxiae and portlandica is restricted to the 
Volgian. 

Rhynchonella rivelensis Loriol 
(PI. 5, figs. 3-6, text-fig. 18) 

1904 Rhynchonella rivelensis Loriol: 282-83, pi- 2 °' fig- 43- 

Emended diagnosis. Small to medium sized Rhynchonella ; subpentagonal in 
outline ; slightly incurved interarea ; large smooth area posteriorly about 10 strong, 
subrounded ribs anteriorly ; crura radulifer. 

Stratigraphical range. Oxfordian — upper transversarium and lower pseudo- 
cordata zones. 

Type specimen. Lectotype, here selected, fig. 43, pi. 27 of Loriol (1904), the 
specimen came from the " Argovien II " of Mont Rivel, France. 

Material. 60 specimens from Alex near Oyonnax (Ain) ; 2 specimens from 
Choux du Dombief (Jura). All the material was collected by Dr. Enay and sub- 
sequently donated to the author. 

Description. External characters. R. rivelensis is subpentagonal in outline and 
is either equally biconvex or with the brachial valve the more inflated. There is a 
low uniplication in the anterior commissure and a low fold on the brachial valve. A 
linguiform extension is only poorly developed. The coarse, simple ribs are only 
developed on the anterior half to one third of the shell ; concentric ornament 
consists of growth lamellae, most clearly seen on the posterior smooth area, and very 
fine growth lines. 

The beak is small, suberect and sharply pointed. Beak ridges are present and 
limit the incurved interarea. The deltidial plates project slightly around the hypo- 
thyridid pedicle opening. 

Dimensions of Figured Specimens. 

length thickness width 

i- 16 cm o-66 cm 122 cm 

1-17 cm 0-77 cm 125 cm 

Internal characters. Pedicle valve. The quadrate delthyrial cavity is bounded by 
long subparallel dental lamellae ; the lateral cavities are well developed. The 
teeth are strong but lateral denticulae are scarcely differentiated. 



JURASSIC RHVNCHONELLIDS 



Brachial valve. There is a long, low median septum ; septalial plates are not 
developed. Inner and outer socket ridges are both present. The clearly marked 
crural bases are sharply pointed ; the radulifer crura become expanded distally. 





10 





1-8 




2-1 





2-5 2-3 

Fig. i8. Transverse serial sections of Rhynchonella rivelensis Loriol. Oxfordian — trans- 
versarium zone. Alex pres Oyannax (Ain), France. (x8). 



FROM NORTHWESTERN EUROPE 45 

Distribution. The only records of the species are from the central French Jura 
(Jura ledonien). 

Occurrence. Nothing is known except that the material supplied by Dr. Enay 
came from a pale coloured marl, as evidenced by the adhering matrix. 

Rhynchonella sp. 
(PI. 5, figs. 1-2) 

Remarks. The species has not been named or described as only four specimens 
are known ; these were donated to the author by Dr. D. A. B. Pearson from the 
Callovian of Kozlowagora in Poland. 

Subfamily IVANOVIELLINAE Makridin, 1964 

Emended diagnosis. Small to medium sized rhynchonellids ; strong, sub- 
angular to rounded ribs ; posterior smooth area ; well marked uniplication and 
linguiform extension often well developed ; septalium present ; crura calcarifer. 

Stratigraphical range. PPliensbachian — middle Oxfordian. 

Genus THURMANNELLA Leidhold 

1920 Seplaliphoria (Thurmannella) Leidhold, pp. 357-62. 
1964 Thurmannella Leidhold; Makridin, pp. 143-44. 
19656 Thurmannella Leidhold; Ager, pp. 609-10. 

Type species. The nominal type species is Terebratula Thurmanni Voltz which 
was designated by Makridin. However, this is regarded by the author as a synonym 
of Terebratula obtrita Defrance and is discussed under the latter species. 

Emended diagnosis. Small to medium sized, subpentagonal, rhynchonellids ; 
ventral valve flattened while the brachial is strongly inflated ; 15-22 simple, sub- 
rounded to subangular ribs which do not arise at the umbones so that there is a 
characteristic but variably developed smooth area posteriorly ; well developed 
uniplication ; linguiform extension usually present but variably developed ; beak 
small, suberect and sharp ; pedicle opening present ; crura calcarifer. 

Stratigraphical range. Oxfordian — mariae to lower transversarium zones. 

Species. T. acuticosta sp. nov. 

T. obtrita (Defrance) (1828, pp. 161-62). 

Distribution. Thurmannella has been recorded from northern France, the 
northern French Jura, the area around Basel and from Yorkshire and Kent. It is 
not recorded from Germany. 

Remarks. A diagnosis only has been given as a full description is given of the 
type species, T. obtrita ; and a generic description would only differ in being broad- 
ened to include T. acuticosta, which differs from T. obtrita in having a very limited 
posterior smooth area, more angular ribs and a more strongly developed uniplication 
in the anterior commissure. 

Thurmannella was first established by Leidhold (1920) as a new subgenus within 
his new genus Septaliphoria. The close relationship thus suggested for Septaliphoria 
and Thurmannella is based on their both possessing a septalium. It would appear 



46 JURASSIC RHYNCHONELLIDS 

that Thurmannella was not officially elevated to generic rank until Makridin did so 
in 1964 ; it is also described as a genus in the Anglo-American " Treatise " and the 
author would certainly agree with its separation from Septaliphoria. 

Thurmannella obtrita (Defrance) 
(PI. 5, figs. 7-9, text-fig. 19) 

1828 Terebratula obtrita Defrance: 161-62. 

1833 Terebratula Thurmanni Voltz; Thirria: 172, 179. 

1842 Terebratula Thurmanni Voltz; Boye: 17, pi. 2, rig. 17. 

1859 Terebratula Thurmanni Voltz; Oppel: 608. 

1 861 Terebratula Thurmanni Etallon: 291, pi. 13, fig. 6. 

1865 Terebratula Thurmanni Ogerien: 656, 661, text-figs. 294-296. 

1871 Terebratula Thurmanni Voltz; Quenstedt: 90-91, pi. 38, figs. 83-87. 

1882 Rhynchonella Thurmanni (Voltz); Haas and Petri: 238-240, pi. 7, figs. 14-17. 

1884 Rhynchonella obtrita (Defrance); Deslongchamps : 337, pi. 38, figs. 1-3. 

1889 Rhynchonella obtrita (Defrance); Haas; 12-13. 

1897 Rhynchonella Thurmanni (Voltz); Loriol: 143-46. 

1900 Rhynchonella Thurmanni (Voltz); Loriol: 135, pi. 6, fig. 53. 

1901 Rhynchonella Thurmanni (Voltz); Loriol: 112, pi. 6, figs. 17-19. 
1904 Rhynchonella Thurmanni (Voltz); Loriol: 283. 

1917 Rhynchonella Thurmanni (Voltz); Rollier: 154-56. 

1917 Rhynchonella silicola Rollier: 155-56. 

1918 Rhynchonelloidea thurmanni (Voltz); Buckman: 38. 

1920 Septaliphoria (Thurmannella) Thurmanni (Voltz); Leidhold: 363-65, pi. 5, fig. 6, pi. 6, 

figs. 10-18. 
1920 Septaliphoria (Thurmannella) obtrita (Defrance); Leidhold: 362-65, pi. 5, fig. 5, pi. 6, figs. 

1-9. 
1964 ^Thurmannella thurmanni (Voltz); Makridin: 135, pi. 6, fig. 53. 

Emended diagnosis. As for genus, except that T. obtrita always has a well 
developed smooth area posteriorly. 

Stratigraphical range. Lower Oxfordian — mariae to lower transversarium 
zones. 

Type specimen. In view of the great confusion in the literature, it has been 
decided that it is necessary to erect a neotype in order to fix the species. The 
specimen proposed was obtained from the " Terrain a Chailles " of Pagny-sur- 
Meurse (Meurthe et Moselle). This is the horizon and area of Defrance's original 
specimens. The specimen is in British Museum (Nat. Hist.) no. BB. 44145. 

Dimensions of neotype. Length 1-40 cm, thickness 1-35 cm, width 172 cm. 

Material. 30 specimens from the " Terrain a Chailles ", Pagny-sur-Meurse ; 
in British Museum (Nat. Hist.) nos. BB. 45720-749. 8 specimens from the " Ter- 
rain a Chailles ", Liesdorf, near Basel ; 100 from the Renggeri Marl of Arc-sur- 
Montenot ; 18 specimens from the Renggeri Marl of Epeugney ; 30 specimens from 
the " ferruginous oolite " of the Verdun area. 

Description. External characters. The smallest specimens are equally bi- 
convex, have a laterally suboval outline and only a very low uniplication. All these 
features become considerably modified with growth and in adult specimens the 
brachial valve is very much the more inflated and generally the uniplication is very 



FROM NORTHWESTERN EUROPE 47 

strongly developed ; is it accompanied by the development of a low fold on the 
brachial valve. However, the development of the uniplication is very variable and 
it may be low and rounded even in large specimens. The outline of adult specimens 
is subpentagonal. 

The beak is suberect in the young specimens but becomes characteristically erect 
in the adult ; it is always relatively small and sharply pointed. At all stages the 
species possesses a well developed, submesothyridid pedicle opening. The beak is 
flanked by sharp beak ridges which clearly delimit an incurved interarea. The 
deltidial plates may be either conjunct or disjunct. 

The 15-20 ribs, which are only present on the anterior half to a third of the shell, 
are low and sub-rounded to sub-angular. The ribs appear to be consistently more 
rounded on the material from the " Terrain a Chailles "; however, this may, at 
least in part, be a result of the different preservation. At the extreme anterior of 
the largest specimens, the ribs sometimes develop a weak trough along their crests. 

Dimensions of Figured Specimens. 

length thickness width 

1 -42 cm 1-17 cm 1-53 cm 

173 cm 170 cm 1-93 cm 

1 40 cm 1-35 cm 172 cm 

Internal characters. The specimen sectioned was largely silicified with a conse- 
quent loss of much fine detail internally. 

Pedicle valve. The subquadrate delthyrial cavity is flanked by strong subparallel 
dental lamellae, which persist, as seen in transverse section, until the teeth have 
been completely inserted in their sockets. The teeth do not appear to be crenulated ; 
lateral denticulae are not developed. 

Brachial valve. A septalium is present, although, judging from the sectioned 
specimen and from the silicified material treated with acid, both the median septum 
and the septalial plates vary considerably in their development. However, some of 
this apparent variation is almost certainly a result of the incomplete silicification. 

Crural bases are well developed ; inner and outer socket ridges are both present. 
The crura appear to conform in general appearance with the " calcarifer " type 
described by Muir-Wood (1936). Distally the cura are blade-like and falciform, as 
can be seen in text-figure 19, however, at their posterior ends they are bent inwards, 
at almost right angles to their distal ends, to form a " flange ". These " flanges " 
are undoubtedly the structures Leidhold (1920) figured and described as " Crural- 
spitze ". This feature differentiates the crura from the true falcifer type as 
developed in Lacunosella, and it seems that it is the diagnostic feature of calcarifer 
crura as defined by Muir-Wood (1936). In her original description Muir-Wood 
described calcarifer crura as consisting " of two flattened, curved, posteriorly con- 
cave laminae which project from the hinge plate into the cavity of the dorsal valve. 
These laminae each unite with a second curved lamina which appears to be suspended 
from it and projects dorsally like a spur. A ventral extension of this second lamina 
terminates in a hook-shaped process, the apex of which is directed medianly ". 



4 8 



JURASSIC RHYNCHONELLIDS 




=7 TO 




2 8 



) ? 1 



3-2 



3 6 



3 4 



Fig. 19. Transverse serial sections of Thurmannellaobtrila (Deivance). Oxfordian. "Terrain 
a Chaillcs ". Pagny-sur-Meurse (Meurthe et Moselle), France. ( X5). 



FROM NORTHWESTERN EUROPE 49 

Distribution. The species occurs quite widely in the lower Oxfordian of the 
northern rim of the Paris Basin, in the northern part of the French Jura and in the 
Swiss Jura around Basel. T. obtrita seems to be co-extensive with the " Terrain a 
Chailles " ; whether it is equally widespread within the Renggeri Marls is not known, 
although the remarks of Loriol (1900) would suggest that it is. 

Occurrence. The species has been collected in quantity from two horizons 
represented by three markedly differing lithologies. The lower horizon is represented 
by the Renggeri Marl, which is a dark coloured calcareous clay. The specimens 
collected showed a very wide range of size and it is presumed that all except the very 
youngest stages are present. It seems reasonable to assume that the Marl was 
deposited in a fairly low energy environment. 

The Renggeri Marl belongs to the mariae zone, while the higher horizons at which 
T. obtrita occurs are in the cordatum and transversarium (lower part) zones. The 
two distinctive lithologies occurring in the upper zones are the stratigraphically 
equivalent " Terrain a Chailles ", a sandy limestone with nodular layers of silica, 
and the self explanatory " oolithe ferrugineuse ". 

The specimens of T. obtrita found in the " Terrain a Chailles " tend to be con- 
siderably larger than those from the Marls, both in average size and in the absolute 
size of the largest specimens ; it is also noticeable that there are fewer of the very 
small individuals. It is suggested that the average size of the specimens within the 
species probably increased through its stratigraphical range and that the larger size 
of the specimens from the " Terrain a Chailles " is emphasized by the higher energy 
environments which is assumed from the lithology ; the smaller shells were probably 
carried away. 

The " Terrain a Chailles " is coarsely bedded and the tops of the bedding planes 
are rather nodular and often show evidence of burrowing activity. T. obtrita is 
found predominantly on these nodular surfaces, where it occurs both singly and in 
groups. Nearly all the specimens are present with complete, articulated shells, 
suggesting that they were not transported far before burial ; this assumption is 
supported by the very delicate nature of the test. 

From the thin shell and the large, obviously functional pedicle opening, it is 
suggested that the individuals may possibly have lived attached to weed and that 
on death they became detached and dropped to the bottom. The groups could be 
either post-mortem accumulations washed together or possibly small colonies which 
were deposited together from the same fronds of weed ; if they in fact adopted that 
mode of life. The " Terrain a Chailles " deposits are thought to have accumulated 
in water no deeper than about 10 fathoms and possibly quite close to the shore line. 

The above remarks on the " Terrain a Chailles " are probably also largely applic- 
able to the " oolithe ferrugineuse ", which is about 7 m thick in the area of Verdun 
where the specimens were collected. The bed is only poorly exposed, largely in old 
shell craters, and collecting from it was relatively difficult. T. obtrita does not seem 
to be so abundant as in the " Terrain a Chailles ", but the specimens are about the 
same size. Associated fauna included terebratulids, irregular echinoids (HCollyrites), 
and lengths of Miller ecrinus. 



50 JURASSIC RHYNCHONELLIDS 

Remarks. The considerable confusion as to the correct nomenclature of the 
species has largely arisen through the inadequacy of the original descriptions. The 
name obtrita was first used in 1828 by Def ranee in the " Dictionaire des Sciences 
Naturelles "; the description was very generalized and while no figure was given, 
he referred to plate 241, fig. 5 in the " Encyclopodie Methodique ". This latter 
figure is not sufficiently clear for specific identification, although the specimen 
represented certainly belongs within the Thurmannella, Rhynchonelloidella, Ivano- 
viella group. While the specimen figured in the " Encyclopodie " is not named or 
geographically located, in 1820 Schlotheim remarked that it was similar to his 
Terebratulites varians ; this fact is also of interest in that the volume of the " Ency- 
clopodie " in which the figure appeared is generally stated to have been published 
in 1827. 

The name Thurmanni first appears in Thirria's (1833) work " Statistique Mineral- 
ogique et Geologique du Departement de la Haute-Saone ", in which the palaeonto- 
logical identifications were given by Voltz. There is no description or figure and the 
name only occurs in the faunal lists for the " argile avec chailles " and " marne 
moyenne avec minerai de fer oolithique ". The first adequate figure of the species is 
by Boye (1842) in his " Importance de l'Etude des Fossiles pour la Reconnaissance 
geologique des Terrains "; again there is no description and the only information 
given is that T. Thurmanni occurs infrequently in " la marne oxfordienne ". Boye 
placed this between the " Chailles " and the " Dalle nacree ", making it almost 
certainly the " Renggeri Marls " of present nomenclature. 

Between 1828 and 1884 the name obtrita does not appear to have occurred in the 
literature, contrasting with thurmanni which was widely used, and the species was 
well figured by several authors under that name. However, in 1884 Deslongchamps 
published a paper in which he redefined the species originally described by Defrance 
(1828), using specimens from the Defrance collection. In this work Deslongchamps 
admitted that the figure in the " Encyclopodie " cited by Defrance was not clear 
but went on to say that virtually all the specimens of Defrance labelled as obtrita 
belonged to the form subsequently named thurmanni. Deslongchamps considered 
some of the Defrance specimens to have come from the " Terrain a Chailles " of 
" Bourgogne or Franche Compte ", and not from near Brussels as stated by Defrance, 
while others came from Lons-le-Saunier, Besancon and Mezieres. He then fully 
described the species, i.e. obtrita, and figured three specimens. However, while it is 
assumed that the figured specimens were from the Defrance collection, this is not 
certain as a copy of the original work is not available at the British Museum. It is 
presumed from contemporary synonymies that Deslongchamps produced this paper, 
with its two plates and their explanations, as a separate publication in 1884. 

When the paper on the Defrance species was published in the " Bulletin de la 
Soci6t6 linni£nne de Normandie " in 1885, for some reason the explanations to the 
plates were omitted. From the remarks of Deslongchamps, there seems little doubt 
as to what Defrance understood by obtrita even though this was not clearly stated in 
his original description. It is presumed that the Defrance collection, which was at 
Caen, was destroyed together with the Deslongchamps collection during the invasion 
of Normandy in 1944. 



FROM NORTHWESTERN EUROPE 51 

Since the Deslongchamps paper, most authors, with the notable exception of Haas 
(1889), have continued to use the name thurmanni. Rollier (1917) suggested that 
the name obtrita should be abandoned and at the same time proposed the restriction 
of the name thurmanni to the smaller forms as typified by the material from the 
renggeri Marl. He placed the larger specimens coming from the " Terrain a 
Chailles " in a new species silicola. To the present author, this arbitrary division 
seems to have little justification, apart from being inadmissible on the grounds of 
priority. 

In setting up his new subgenus Thurmannella, Leidhold (1920) included only two 
species, namely T. obtrita and T. thurmanni, nominating neither as type of the genus. 
These two ' species ' were differentiated on the grounds that the valves of T. obtrita 
were less inflated and that the uniplication was less developed than in T. thurmanni ; 
however, as they occur together at the same horizon and localities and as the features 
mentioned above show complete gradation between the ' end members ', it is here 
considered that Leidhold was in fact dealing with one species. The synonymy given 
by Leidhold for his two species is interesting in that he placed T. thurmanni of Boye 
in synonymy with T. obtrita of Def ranee, so separating it from T. thurmanni of 
Voltz ; the grounds on which this decision was made are not made clear. 

It is concluded that the name obtrita of Def ranee has priority over thurmanni of 
Voltz and that T. obtrita should be regarded as the type species of Thurmannella. 



Thurmannella acuticosta sp. n. 

(PI. 5, figs, io-ii, text-fig. 20) 

1878 Rhynchonella varians var. Thurmanni (Voltz); Davidson: 215-16, pi. 28, figs. 14-16. 
1933 Rhynchonelloidea thurmanni Arkell: 432. 

Name. Latin, acutus-a-um — sharp, pointed ; costa-ae — a rib ; the species has 
very sharp ribs. 

Diagnosis. Medium sized Thurmannella ; about 17-18 relatively coarse, steep, 
subangular ribs ; possesses only very limited posterior smooth area ; crura cal- 
carifer. 

Stratigraphical range. Lower Oxfordian — mariae zone and lower cordatum 
zones. 

Type specimen. Holotype selected is the specimen figured Davidson (1878) pi. 
28, fig. 16. 

Dimensions of holotype. Length i-6i cm, thickness 1-46 cm, width 1-74 cm. 

Description. External characters. The shell outline is subpentagonal to 
weakly trilobate, the latter condition resulting from the very strongly developed 
uniplication shown by some specimens. The uniplication in the anterior commissure 
is emphasized by the strong, steep nature of the subangular ribs. No concentric 
ornament has been observed. 

The beak is sharp and suberect to erect ; the erect condition is much less common 
than in T. obtrita and, in comparison with that species, the interarea is smaller and 
less incurved. An oval, submesothyridid pedicle opening is present ; the details of 



5-: 



JURASSIC RHYNCHONELLIDS 



the deltidial plates have not been satisfactorily observed. The smooth area on the 
brachial valve is restricted to not more than the posterior quarter of the valve. 



Dimensions of Figured Specimens. 

length thickness 

1-43 cm iio cm 

i'44 cm I-I2 cm 



width 
i-6o cm 
1-67 cm 



Internal characters. The preservation of the available specimens is very poor and 
it was found that the internal features of the first specimen were completely oblitera- 
ted by partial silicification. A second specimen was sectioned and the structures, 




Fig. 20. Transverse serial sections of Thurmannella acuticosta sp. nov. Oxfordian. Lower 
Calcareous Grit. Filey Brigg, Yorks. (X5). 



FROM NORTHWESTERN EUROPE 53 

although still only poorly preserved, appeared to show no notable differences from 
those described above for T. obtrita. 

Distribution. T. acuticosta is one of the most common species in the Lower 
Calcareous Grit of Yorkshire but, according to Arkell (1933), it has only been 
recorded from two localities in the south of England, namely, at Catcombe, near 
Lyneham, and in one of the Kent borings. 

Occurrence. The Lower Calcareous Grit appears to be very similar to the 

' Terrain a Chailles ", as developed in the cor datum and lower transversarium zones 

of northern France, with the exception of the abundance of " Spongia paradoxica " 

(=the trace fossil Thalassinoides teste Ager), as seen at Filey Brigg. It would 

appear that the environment represented is again of the nearshore, high energy type. 

Subfamily ACANTHOTHYRIDINAE Schuchert, 1913 
Genus ACANTHOTHIRIS Orbigny 

1850 Acanthothiris Orbigny, 323. 

1875 Acanthothyris Orbigny: Paetel, 1. 

1877 Acanthothiris Dall, 11. 

1886 Gruppe der Spinosen Rothpletz, 93. 

1889 Acanthothyris Orbigny: Buckman and Walker, 41. 

1894 Acanthothyris Orbigny: Hall and Clarke, 836. 

1914 Acanthothiris Orbigny: Buckman, 2. 

1917 Acanthothyris Orbigny: Rollier, 74. 

1918 Acanthothiris Orbigny: Buckman, 70. 

1929 Acanthothiris Orbigny: Schuchert and Le Vene, 26. 

1936 Acanthothyris Orbigny: Muir-Wood, 27. 

i960 Acanthothiris Orbigny: Ager, 157. 

i960 Acanthothyris Orbigny: Makridin, 256. 

1963 Acanthothiris Orbigny: Seifert, 168. 

19656 Acanthothiris Orbigny: Ager, H611. 

Type species. Terebratulites spinosus Schlotheim, designated by Rollier (1917). 
This is regarded by the author as a synonym of Anomia spinosa Linnaeus (1767) and 
is discussed below. 

Emended diagnosis. Small to medium sized sub-pentagonal or laterally oval 
rhynchonellids ; biconvex ; uniplicate with low dorsal fold ; many well developed 
ribs, several of which bifurcate ; test covered by hollow spines which arise from the 
crests of the ribs ; beak small and erect or slightly incurved ; well developed groove 
(" pedicle trough ") running from posterior of pedicle opening to tip of beak ; fairly 
strong median septum present ; crura radulifer. 

Stratigraphical range. ? Aalenian, Bajocian — Bathonian/basal Callovian. 

Description. External characters. The larger acanthothirids tend to be more 
globose than the smaller ; the brachial valve is more inflated than the pedicle. 
The many well marked, rather shallow, sub-rounded ribs tend to bifurcate laterally 
and from their crests arise the long, slender, hollow spines which cover the test. 
There is a slight fold on the dorsal valve and a rounded uniplication in the anterior 
commissure. The beak is relatively small and sufficiently incurved to obscure the 



54 JURASSIC RHYNCHONELLIDS 

details of the pedicle opening and deltidial plates. Beak ridges and interarea are 
present but rather variable in development. The young forms tend to be sub- 
trigonal to subcircular. 

Internal characters. Pedicle valve. The delthyrial cavity is quadrate in trans- 
verse section, but in the initial posterior sections there is a low dorsally directed 
ridge. This results from the presence of a groove which is situated on the antero- 
dorsal side of the beak and runs from the pedicle opening to the tip of the beak. 
This structure appears to be a pedicle collar as originally defined by Thomson (1927) ; 
it is not, however, the structure defined as a pedicle collar in the " Treatise " (1965), 
and is here referred to as the " pedicle trough ". 

The lateral cavities are relatively narrow and the slender dental lamellae bounding 
them are either subparallel or slightly divergent ventrally. Teeth and sockets are 
both crenulated. 

Brachial valve. A median septum is quite well developed. A septalium is not 
present but the median structure, dorsal of the ventrally arched hinge plates, seen in 
A. spinosa and A. cf. A. costata could possibly have resulted from the fusion of 
septalial plates. Crural bases are hardly differentiated. The crura are of the 
radulifer type and tend to be oval or flattened in the plane of articulation, as seen in 
transverse section. Muscle scars have not been observed and even Buckman (1918) 
commented that they were obscure in both valves. 

Species. As well as A. spinosa and A. cf. A. costata, here described, the literature 
abounds with descriptions and names of spinose rhynchonellids of this age. The 
following nominal species are provisionally attributed to this genus : 

A. balinensis Rollier (1917, p. 77) 

A. broughensis Muir-Wood (1952, pi. 123, p. v, figs. 11, 13-15) 

A. crossi (Walker) (1869, p. 215) 

A. costata Orbigny (1850, p. 286) 

A. doultingensis (Richardson and Walker) (1907, p. 426, pi. xxvii, fig. 2) 

A. elargata Seifert (1963, p. 170, pi. x, fig. 8) 

A. globosa Buckman (1918, p. 918, pi. xix, fig. 25) 

A. inflata (Quenstedt) (1868, p. 112, pi. xxxix, figs. 46-47) 

A. midfordensis (Richardson & Walker) (1907, p. 427, pi. xxvii, fig. 1) 

A. multispinosa Seifert (1963, p. 172, pi. x, fig. 12) 

A. oligacantha (Branco) (1879, p. 127, pi. vi, figs. 6-7) 

A. paucispina Buckman and Walker (1889, pp. 52-53) 

A . power stockensis Buckman and Walker (1889, p. 52) 

A. radwanowicensis Rollier (1917, p. 77) 

A. sentosa (Quenstedt) (1868, p. 113, pi. xxxix, figs. 55-56) 

A. sinuata (Quenstedt) (1868, p. 114, pi. xxxix, fig. 59) 

A. subglobosa Seifert (1963, p. 171, pi. x, fig. 11) 

A. tenuispina (Waagen) 1867, p. 640, pi. xxxii, fig. 6). 

Distribution. Acanthothiris is a common genus occurring widely throughout 
the Middle Jurassic rocks of Europe, including those of England, Normandy, the 
French, Swiss and German Jura and Poland. 



FROM NORTHWESTERN EUROPE 55 

Occurrence. The genus has been collected by the author only from coarse, 
bioclastic limestone. 

Remarks. On account of their distinctive appearance members of Acanthothiris 
have been described and figured on numerous occasions, usually under the name of 
Rhynchonella, Terebratula or Acanthothiris spinosa. The author has not attempted 
to disentangle the nomenclature and synonymy of these numerous descriptions and 
figures, but has here only attempted to define and clarify the type species together 
with A . cf . costata, which happened to be a very common fossil in the part of the 
southern French Jura mapped by members of Imperial College. The latter species 
was investigated in order to obtain a wider knowledge of the variation of internal 
structures of the genus as these are not adequately described in the literature. 



Acanthothiris spinosa (Linnaeus) 
(PI. 5, figs. 12-17, text-figs. 21-22) 

1767 Anomia spinosa Linnaeus: 1154. 

1768 (concha) anomia ventricosa, striata echinata, d'Annone in Knorr and Walch: 90, pi. B4, 

fig- 4- 

1779 Anomia Walcott: 21, fig. 31. 

1813 Terebratulites spinosus Schlotheim : 73. 

1813 Anomia spinosa Linnaeus; Townshend: 372. 

1 81 7 Terebratula spinosa Smith: 108. 

1820 Terebratulites spinosus Schlotheim: 269. 

1825 ^Terebratula aspera Koenig: fig. 219. 

1829 Terebratula spinosa Townshend and Smith; Phillips: 123, pi. 9, fig. 18. 

1833 Terebratula spinosa Buch: 58-59. 

1838 Terebratula spinosa Buch: 161-62, pi. 6, fig. 4. 

1850 Hemithiris spinosa Orbigny: 268. 

1850 Acanthothiris spinosa Orbigny: 323. 

1851 Rhynchonella spinosa (Schlotheim); Davidson: 71-73, pi. 15, figs. 15-20. 

1857 Rhynchonella spinosa (Schlotheim); Deslongchamps: 355-56, pi. 5, fig. 1. 

1858 Rhynchonella spinosa (Schlotheim); Davidson: 222. 
1882 Rhynchonella spinosa (Schlotheim); Buckman: 40-41. 

1889 Acanthothyris spinosa (Linnaeus); Buckman and Walker: 12-14. 

1900 Acanthothyris spinosa (Schlotheim); Greppin: 178-80, pi. 19, fig. 3. 

1917 Acanthothyris spinosa (Schlotheim); Rollier: 76-77. 

1918 Acanthothiris spinosa (Linnaeus); Buckman: 70-72, pi. 19, fig. 26. 
1936 Acanthothiris spinosa (Linnaeus); Muir-Wood: 28. 

1936 Acanthothyris spinosa (Schlotheim); Arcelin and Roche: 54-58, pi. 1, figs. 1-18, pi. 12, 

fig. 2. 
1959 Acanthothyris spinosa (Schlotheim); Henry: 119-21. 
1966 Acanthothiris spinosa (Schlotheim); Almeras: 277-79, pi. 17, figs. 1-3. 

Emended diagnosis. Medium to large sized Acanthothiris ; relatively globose ; 
outline subpentagonal ; brachial valve considerably more inflated than the pedicle ; 
" pedicle trough " present ; well marked uniplication ; 30-35 ribs on each valve ; 
crura radulifer. 

Type specimen. Neotype, here proposed, is the specimen in the collection of the 
B.M.(N.H.) from the Inferior Oolite of Rodborough Hill numbered B4030. The 



56 JURASSIC RHYNCHONELLIDS 

exact horizon is not given but it is probably from the Upper Trigonia Grit, i.e. the 
parkinsoni zone of the Bajocian. 

Stratigraphical range. Upper Bajocian to Upper Bathonian/Lower Callovian. 

Dimensions of neotype. Length 2-14 cm, thickness 171 cm, width 2-37 cm. 

Material. 30 specimens from the collection of the B.M.(N.H.) ; 23 specimens 
from Normandy and 9 from Switzerland in the collection of the author in the British 
Museum (Nat. Hist.) nos. BB. 45682-89. 

Description. External characters. Medium to large sized Acanthothiris which 
is characteristically fairly globose. In early growth stages the two valves are of 
almost equal convexity but, in adult forms, the brachial valve is much the more 
inflated. Young forms have a subcircular to subtrigonal outline, which becomes 
modified to the subpentagonal shape of the adult. The width is usually, but not 
invariably, greater than the length. There is a marked, rounded uniplication in the 
anterior commissure, accompanied by a low fold on the brachial valve. 

In early growth stages, the beak tends to be small, acute and erect ; it becomes 
more incurved in the adult and often obscures the pedicle opening and deltidial 
plates. The deltidial plates, as seen in the serial sections, are conjunct ; the pedicle 
opening is submesothyridid. Beak ridges are quite well developed and limit the 
incurved interarea. 

The test ornament consists of about 30 to 35 subangular to subrounded ribs, as 
counted at the anterior commissure ; these ribs may bifurcate once or twice, rarely 
more, and occasional ribs are inserted by intercalation. The long, slender spines 
which form the most distinctive feature of the genus arise from expanded bases 
situated along the crests of the ribs. Spines of up to 1 cm have been observed still 
attached to the test ; however, in most specimens they are indicated only by the 
presence of low spine bases. Although they are hollow, they do not seem to have had 
any direct connection with the mantle. The other ornament consists of rather weak 
growth lines, which are usually only seen in the inter-rib troughs of well preserved 
material, and occasional growth lamellae. 

Davidson (1858) suggested that, " when alive, R. spinosa, was more or less coloured 
with red ; at least, we have seen specimens in which the spines had preserved that 
colour ; this is also the opinion of M. Deslongchamps ". This observation was 
probably founded largely on material from quarries near Caen in Normandy and, 
while true, omits to point out that large numbers of other organisms are also so 
coloured and this would appear to be due to some local factor rather than to original 
colouration. Dr. Rioult, (personal communication, 1964) suggested that this 
colouration is due to algae. 

Dimensions of Figured Adult Specimens. 



length 






thickness 


width 


2-28 cm 






178 cm 


2-37 cm 


2-14 cm 






171 cm 


2-37 cm 


2-42 cm 






2-05 cm 


2-88 cm 


Internal characters. 


Pedicle 


valve 


. The develop 


<ment of the de^ 



showing the crumpling together characteristic of the species, is clearly illustrated in 



FROM NORTHWESTERN EUROPE 





"S-w. 



31 




3 4 




3-7 




«N 




^ 



5-9 



50 



4-1 



Fig. 2i. Transverse serial sections of Acanthothiris spinosa (Linnaeus). Bajocian. Upper 
Trigonia Grit. Wooton-under-Edge, Glos. (X4). 



58 J URASSIC RHYNCHONELLIDS 

the serial sections. The pedicle opening can be seen at the base of the pedicle 
groove ; the difference in appearance of this structure, as compared with A. cf. 
costata results from the greater incurvature of the beak in A . spinosa. 

Brachial valve. See generic description. The shape of the crura seems to be a 
useful criterion for differentiating the species from A. cf. costata. 

Distribution. The species has been definitely recorded from the Cotswolds, 
Normandy, northern France, around Basel in Switzerland and from Monsard 
(Saone et Loire). The name spinosa has been used frequently in the past for any 
spinose rhynchonellid and consequently from the literature, it is very difficult to 
ascertain the true distribution of A. spinosa s.s. 

Occurrence. The species is considered to have lived under conditions similar to 
those described for A . cf . costata. 

Remarks. The rhynchonellid now known as Acanthothiris spinosa s.s. was first 
figured by Knorr and Walch (1768) with the description, " concha anomia, ventri- 
cosa, striata echinata ", which was given by Professor d'Annone. A specimen 
regarded by Rollier (1917) as the type of Knorr and Walch is preserved in the 
Naturhistorisches Museum of Basel ; however, as Rollier pointed out, this specimen 
differs markedly from the original figure in such features as the number of ribs, of 
which the specimen has 36-38 on the pedicle valve rather than the 26 figured. This 
discrepancy is such that Dr. Gasche, of the Museum, is of the opinion (personal 
communication 1965), that this is not the original of the Knorr and Walch figure and 
the present author would concur with this observation. 

Linnaeus (1767)* described Anomia spinosa, stating that it had long spines and 
that it came from England. Although, unfortunately, the original is not preserved 
in the Linnaeus Collection, at present in the B.M.(N.H.), it seems reasonable to 
assume that the A. spinosa referred to by Linnaeus is the very common Inferior 
Oolite species of Acanthothiris. This was the view taken by Muir-Wood in her 1936 
monograph. Therefore, it has been decided to define the species with a neotype 
from the Bajocian of the Cotswolds. 

The confusion over the authorship of the species has arisen mainly as a result of 
Schlotheim (1813) mentioning a Terebratulites spinosus, which he did not figure or 
describe, but merely referred to the figure of Knorr and Walch. As this figure was 
not named in the correct binomial form and as it is regarded by the present author as 
conspecific with Anomia spinosa of Linnaeus, the Schlotheim name must be regarded 
as a synonym and the species attributed to Linnaeus. 

In 1889 Buckman and Walker fully discussed the earlier literature and redescribed 
all the known British spinose rhynchonellids. Rollier added to the confusion by 
setting up many new species based purely on published figures, many of which were 
wholly inadequate in the first place. Buckman (1918) redefined the genus and 

* This is the first edition in which spinosa is described and at the end of the remarks on habitat, which 
mention that the species occurs in England, is written " Solander ". Daniel Carl Solander was an 
associate of Linnaeus but it is not clear whether, by placing Solander's name after the remarks Linnaeus 
was suggesting that he was the author of the species or whether he was merely quoting Solander in 
stating that the species occurred in England. All subsequent authors who have referred to the " Sys- 
tema Natura " have attributed the name spinosa to Linnaeus. 



FROM NORTHWESTERN EUROPE 



59 




2-9 



34 



4 4 






54 52 50 

Fig. 22. Transverse serial sections of Acanthothiris spinosa (Linnaeus). Bathonian. " Var- 
ians Schichten ", Liestal, Basel, Switzerland. (X4). 



6o JURASSIC RHYNCHONELLIDS 

although no specific description of A. spinosa was given, he figured a specimen of 
A. spinosa s.s. from the Inferior Oolite of Midford near Bath. 

There was no further discussion of the species until 1936 when Arcelin and Roche 
attempted to disentangle the synonymy, but failed to take account of the literature 
prior to Schlotheim. They described and figured a beautiful series of silicified 
specimens from the Bajocian of Monsard (Saone et Loire) ; these specimens have 
been examined in the collection of the University of Lyon and appear to belong to 
A. spinosa s.s., although it is difficult to compare the internal structures with those 
of the sectioned material. Also in 1936 Muir-Wood discussed the type of the genus, 
concluding it to be the species referred to by Linnaeus and that this was the " com- 
mon Inferior Oolite species " ; in doing this, however, she appears to have overlooked 
the earlier designation of Rollier. She also stated that she considered the spinosa 
of Linnaeus to be congeneric but not conspecific with the form described by Knorr 
and Walch from Muttenz, as specimens from that area were " larger and more 
globose ". 

The area of Muttenz now being an industrial suburb of Basel, it is not possible to 
obtain exactly topotypic material ; however, within a few kilometres there are good 
exposures of the Upper Bathonian Varians-Schichten from which the original is 
presumed to have come and the author visited one such exposure at Liestal, collecting 
a good sample of specimens. From the examination of this material, the author 
believes that it should be considered as conspecific with the English species from the 
Inferior Oolite ; specimens from both areas have been sectioned, see text-figs. 21, 22 
and are identical internally. 

Acanthothiris cf. costata Orbigny 
(PI. 6, figs. 2-4, text-fig. 23) 

1850 Hemithiris costata Orbigny: 286-87. 

1850 Acanthothiris costata Orbigny: 323. 

1910 Hemithiris costata Orbigny; Thevenin: 100, pi. 20, figs. 18-20. 

1917 Acanthothyris costata (Orbigny); Rollier: 74. 

191 8 Acanthothiris costata Orbigny; Buckman: 71. 

Emended diagnosis. Medium sized Acanthothiris ; width always greater than 
length ; outline sub-oval ; 25-30 low, rounded ribs ; beak small and tightly 
pressed to the brachial valve ; crura radulifer. 

Stratigraphical range. d'Orbigny's material came from the " Bajocien ", 
while the specimens here described are of Upper Bathonian/Lower Callovian age. 

Material. Numerous specimens collected by the author and various members 
of Imperial College from the southern French Jura. 20 specimens in the author's 
collection in British Museum (Nat. Hist.) nos. BB. 45690-45709. 

Description. External characters. The width is always greater than the length, 
resulting in a laterally suboval outline. The beak is very small and closely ad- 
pressed to the brachial valve ; no pedicle opening or deltidial plates are visible 
externally in the adult specimens. Beak ridges and interarea are both poorly 
developed. Adult specimens have between 25 and 30 ribs, as counted at the anterior 



FROM NORTHWESTERN EUROPE 61 

commissure. These ribs are rather rounded in cross section and frequently bifurcate 
once or twice over their length ; ribs are also occasionally inserted by intercalation. 
The ornament also includes growth lines, but these are not strongly developed and 
are only seen on well preserved material. Long, slender spines arise from the crests 
of the ribs and are fairly evenly distributed over the test ; the hollow nature of these 
is clearly seen in the serial sections. There is a low rounded uniplication in the 
anterior commissure ; however, any accompanying fold in the brachial valve is only 
poorly developed. 

Dimensions of Figured Specimens. 

length thickness width 

1-96 cm 1 45 cm 2 57 cm 

1-87 cm 1-38 cm 2-50 cm 

1-84 cm 1-15 cm 2-30 cm 

Internal characters. Pedicle valve. The deltidial plates, as seen in serial section, 
are very weakly developed, as compared with A. spinosa, and do not meet and 
become crumpled together. The nature of the " pedicle trough " is clearly seen 
but from the incurvature of the valve it could obviously have been functional only 
during the early stages of ontogeny as proved by the inflation of the brachial valve 
and the incurvature of the beak which must have lead to atrophy of the pedicle in 
adults. The dental lamellae diverge slightly ventrally and limit a subquadrate 
delthyrial cavity. The teeth are strong and crenulated. 

Brachial valve. The median septum is quite strongly developed. There is no 
septalium as the septalial plates, although present, are very weak. Crural bases are 
not differentiated. The crura are radulifer and curve distally toward the pedicle 
valve ; at their proximal ends they are rather flattened in cross section and lie in the 
plane of articulation. 

Distribution. The material on which Orbigny based his original description 
came from Gueret (Sarthe) ; all the specimens described by the author came from 
the southern French Jura (Ain). 

Occurrence. Details of the type locality for A. costata are not known. The 
Jura material was collected from the Middle Chanaz Beds (Ager & Evamy, 1963) 
from the area around Belley (Ain). Dr. Evamy gives the following description of 
the lithology in his thesis: " irregularly bedded, poorly sorted calc-arenites, charac- 
terized ... by their content of ferruginous ooliths The lower and principal 

unit . . . consists of irregularly bedded brown and grey limestones and similarly 
coloured shales ". He also notes, " The member (Middle Chanaz Beds) may also 
be recognized in the field by the abundance of Acanthothiris ". 

The acanthothirids occurred as clusters or " nests " in the limestones and concen- 
trations of them were also noticed in the shales. In both situations the forms were 
mainly adult with valves intact and, in the majority of cases, with the beaks pointing 
downwards. It would seem that most of the specimens were not far removed from 
their life position ; however, some movement or sorting is indicated by the notice- 
able lack of juvenile forms and the similar lack of in situ spines. It should be noted 



62 



JURASSIC RHYNCHONELLIDS 



that while many of the specimens were in " nests " or associations, either vertical or 
horizontal, many others occurred singly and scattered through the rock unit. The 
above mentioned " nests " have been referred to by Ager & Evamy (1963) as seeming, 
" to be true colonial associations ". 




38 





Fig. 23. Transverse serial sections of Acanthothiris cf. costata Orbigny. Bathonian/Cal- 
lovian. Middle Chanaz Beds. Pugieu (Ain), France. (X4). 






FROM NORTHWESTERN EUROPE 63 

The evidence of the lithology is such as to suggest a rather high energy environ- 
ment. Given that the latter deduction is correct, it seems to the author that the 
spines would have been a useful anchoring device and it is considered that Rudwick 
(1964) under-estimated this function in proposing a sensory role for them. 

Remarks. A plaster cast of the specimen from the Orbigny collection figured by 
Thevenin (1910) has been examined, through the kindness of Dr. Fischer of the 
Jardin des Plantes (Mus^e d'histoire Naturelle), and externally this closely resembles 
the material collected from the southern French Jura. However, a conspecific 
identification has not been given as no topotype material has been sectioned and 
hence no comparison of the internal structures has been possible. The main 
difference between the Orbigny specimen and those described from the Jura lies in 
the rib count ; that of the former being 20 while that of the latter is between 25 and 
30. As the other two specimens of A. costata in the Orbigny collection are only 
poorly preserved (Dr. D. V. Ager, personal communication 1965), it is not considered 
reasonable to separate the Jura form as a new species merely on the basis of the 
slight difference in the number of ribs without an investigation of topotype material 
of the Orbigny species to ascertain its variation and internal structure. 



Genus ACANTHORHYNCHIA Buckman 

1914 Acanthorhynchia Buckman: 2, (nomen nudem). 
1918 Acanthorhynchia Buckman: 69. 
i960 Acanthorhynchia Buckman; Makridin: 256. 
19656 Acanthorhynchia Buckman: Ager: H611. 

Type species. Acanthothyris panacanthina Buckman and Walker, by original 
designation. 

Emended diagnosis. Small to medium sized rhynchonellids ; transversely oval 
or sub-triangular in outline ; thin test covered by numerous fine spine-covered ribs ; 
anterior commissure rectimarginate or with slight uniplication ; deltidial plates 
absent, or reduced and disjunct ; crura calcarifer or radulifer. 

Strati graphical range. Upper Bajocian — Upper Volgian/Neocomian. 

Description. External characters. The genus contains a series of small to 
medium sized rhynchonellids. The smaller species tend to be subtriangular in out- 
line, equally biconvex and generally flatter, while the larger species are transversely 
oval in outline, more globose and with the brachial valve more inflated than the 
pedicle. 

The anterior commissure may have a slight uniplication in the larger forms, but in 
the smaller it is usually straight. The beak is relatively small and suberect or erect ; 
deltidial plates are absent in the smaller forms and of a reduced, disjunct variety in 
the larger. Beak ridges and interarea are never well developed. There are typically 
one or two well developed growth lamellae, but the finer growth lines have not been 
observed. The test is covered with many fine, bifurcating ribs, from the crests of 
which are developed the characteristic, numerous fine spines which project at a low 
angle. 



64 JURASSIC RHYNCHONELLIDS 

Internal characters. Pedicle valve. The dental lamellae are apparently only 
developed in the larger species and, when they are not developed, the teeth tend to be 
inserted at a very shallow angle. 

Brachial valve. A low median septum may be present but septalial plates and 
a septalium are never developed. The ends of the inner hinge plates show a dis- 
tinctive thickening and in one species, A. (Acanthorhynchia) spinulosa, this results in 
a structure comparable with a cardinal process. The crura, especially in A. (Acan- 
thorhynchia) panacanthina, the only really well preserved species examined, appear 
to closely resemble the form described by Muir-Wood (1935) as calcarifer. How- 
ever, in other species they seem to more closely resemble the radulifer type. 

Species. The following nominal species are provisionally recognized as belonging 
to the genus ; a sub-generic designation is given where possible : 
A. davidsoni (Rollier) (1917, p. 77) 
A. (Echinirhynchia) dealbata (Rollier) (1917, p. 80) 
A. (Echinirhynchia) fileyensis (Buckman & Walker) (1889, p. 56) 
A. (Echinirhynchia) impressata (Rollier) (1917, p. 79) 
A. (Echinirhynchia) lorioli (Rollier) (1917, p. 79) 

A. (Acanthorhynchia) multistriata (Kitchin) (1900, pp. 75-77) pi. xiv, figs. 10, n 
A. (Acanthorhynchia) myriacantha (Deslongchamps) (1859, P- 2 5 x > pl> i v > n 8- I2 ) 
A. nikitensis Gerasimov (1955, p. 216, pi. xliii, fig. 5) 
A. (Acanthorhynchia) panacanthina (Buckman & Walker) (1889, p. 53) 
A. (Acanthorhynchia) regans (Szajnocha) (1879, p. 222, pi. vi, figs. 1, 2) 
A. (Echinirhynchia) senticosa (Schlotheim) (1820, p. 268) 
A. (Echinirhynchia) silicea (Quenstedt) (1871, p. 115, pi. xxxix, fig. 66) 
A. (Acanthorhynchia) spinulosa (Oppel) (1858, p. 608) 
A. (Acanthorhynchia) vilsensis sp. nov. 
A. zieteni (Rollier) (1917, p. 75). 

Distribution. Although rare, the genus is very widespread and occurs through- 
out the French and German Jura, and in Normandy, Poland and England. It has 
also been recorded from the Kutch. 

Remarks and occurrence. The genus was founded by Buckman in 1918, after 
being mentioned by him in 1914. In his original description, Buckman included 
most of the species here retained in the genus, together with a few now considered as 
belonging to Acanthothiris or to be of uncertain affinity. Although distinctive, all 
the included species are rather rare and this has necessitated the use of much 
borrowed material in order to attempt a complete revision of the genus. 

The genus Acanthorhynchia, as here defined, includes two distinct groups, based 
on external morphology, which overlap each other in time and, to a lesser extent, 
geographical area. It is here proposed to designate these groups as sub-genera, 
namely Acanthorhynchia s.s. and Echinirhynchia sub-gen. nov. 

As stated above, all the species are rare and details of the source rock and supposed 
environment are only rarely given in the literature. However, the fact that A. 
(Acanthorhynchia) panacanthina comes from the Inferior Oolite and A. (Acanthor- 
hynchia) rogans comes from the Balin Oolite suggests that the sub-genus Acanthor- 



FROM NORTHWESTERN EUROPE 65 

hynchia s.s. lived in fairly high energy environments. The specimens of Echinir- 
hynchia collected from the French Jura came from very fine-grained sediments, 
either clays or calclutites, while the material recorded from Germany is also either 
from clays e.g. Terebratula senticosa impressae Quenstedt, or from the sponge reef 
facies e.g. A. (Echinir hynchia) senticosa (Schlotheim). The occurrence of the sub- 
genus in these two markedly different facies may possibly indicate that its mode of 
life was planktonic or that it lived attached to floating weed ; this hypothesis was 
suggested by Ager (1965a). 

Sub-genus ACANTHORHYNCHIA s.s. 

Type species. Acanthothyris panacanthina Buckman and Walker. 

Diagnosis. Medium to large species of Acanthorhynchia with a transversely oval 
to subrounded outline ; either equally biconvex or with a more inflated brachial 
valve ; deltidial plates and dental lamellae may be present. 

Stratigraphical range. Upper Bajocian — Upper Volgian/Neocomian. 

Distribution. The sub-genus has a wide range both geographically and chrono- 
logically ; however, the consistent external and internal features of the included 
species suggest a continuous evolutionary lineage. The extent of the geographical 
range is given under the species description. Apart from the species described A. 
(Acanthorhynchia) multistriata (Kitchin) is recorded from the " Oomia Group " of 
the Kutch and A. (Acanthorhynchia) rogans (Szajnocha) occurs in the " Balin Oolite " 
of Poland. 

A. (Acanthorhynchia) panacanathina (Buckman and Walker) 
(PI. 6, figs. 5-7, text-fig. 24) 

1851 Rhynchonella senticosa (Buch) ; Davidson: 73-74, pi. 15, fig. 21. 

1878 Rhynchonella senticosa (Buch); Davidson: 223. 

1878 Rhynchonella senticosa (Buch); Davidson: 85, pi. 4, fig. 20. 

1882 Rhynchonella senticosa (Buch); Buckman: 41-42. 

1889 Acanthothyris panacanathina Buckman and Walker: 53-54-[=^4. Davidsoni Rollier 1917]. 

1918 Acanthorhynchia panacanathina Buckman and Walker; Buckman: 69-70, pi. 19, fig. 23. 
[=A. Davidsoni Rollier 191 7]. 

Emended diagnosis. Large Acanthorhynchia which ranges in size up to 3-05 cm 
long, 2-20 cm thick and 3-53 cm wide. Biconvex and transversely suboval in out- 
line ; beak suberect ; deltidial plates narrow, trigonal and disjunct ; anterior 
commissure sinuate ; crura calcarifer. 

Stratigraphical range. Bajocian — probably restricted to the parkinsoni zone. 

Type specimen. In their synonymy Buckman and Walker (1889) mention the 
specimen figured by Davidson (1851) on pi. 15, fig. 21 and it seems reasonable to 
take this as the type of the species. B 12082 (B.M. (N.H.)) 

Description. External characters. This is the largest known species of Acan- 
thorhynchia. The test is covered with fine ribs ; Buckman and Walker cite a speci- 
men as having about a hundred ribs on each valve, with forty spines on each rib. 
This statement seems to be essentially correct, although the fineness of the features 
concerned is such as to preclude accurate counting except on well preserved and 



66 



JURASSIC RHYNCHONELLIDS 






/ 

\ 



/ 



/ 



S 



/ 



\ 



\ 



\ 



\ 



\ 



\ 



50 



Fig. 24. Transverse serial sections of A. (Acanthorhynchia) panacanthina (Buckman & 
Walker). Topotype. Bajocian. Burton Bradstock, Dorset. Specimen donated by the 
B.M.(N.H.). (x 4 ). 



FROM NORTHWESTERN EUROPE 67 

carefully prepared material. The ribs maintain a consistent spacing over the 
surface by constant dichotomy. 

The shell is biconvex with the brachial valve the more inflated of the two. The 
beak is small and suberect. Disjunct trigonal deltidial plates flank the elliptical 
submesothyridid pedicle opening. Typically there are one or two well developed 
growth lamellae, although the " very numerous fine growth lines ", noted by 
Buckman (1918) have not been observed on the material studied. The shell being 
very thin and delicate, short, slightly diverging dental lamellae and a short median 
septum are frequently seen, in pedicle and brachial valves respectively, of worn 
specimens. 

Dimensions of Figured Specimens. 



length 


thickness 


width 


1-40 cm 


o-68 cm 


1-55 cm 


3-05 cm 


220 cm 


3-53 cm 


2-46 cm 


173 cm 


2-66 cm 



Internal characters. Pedicle valve. The deltidial plates are only poorly devel- 
oped, as seen externally, but in serial section they are seen to line the subquadrate 
delthyrial cavity laterally. Near the umbo, sections i-o to i-2 in text-fig. 24 are 
seen two thin concave lamellae forming a weak pedicle collar. The dental lamellae 
are quite strong and parallel to one another, not breaking away from the ventral 
shell wall, as seen in serial section, until the teeth are fully inserted in their sockets ; 
the lateral cavities are always relatively small. 

Brachial valve. There is a long, low median septum. Septalial plates are not 
developed but there is a very distinctive thickening of the inner ends of the hinge 
plates, as shown in plate 12. This thickening possibly served for anchoring the 
didductor muscles. Lateral denticulae are present but, in common with the inner 
and outer socket ridges, are not well differentiated. The crura seem to be quite 
close to the shape described by Muir-Wood (1934) as calcarifer. 

Distribution. The author has not collected personally any material of this 
species and, therefore, it seems reasonable to quote the distribution given by 
Buckman and Walker (1889, p. 54) in their original description of the species : 
' This species appears to be confined to the district south of the Mendip Hills. It is 
a rare fossil, and it occurs only in the parkinsoni zone. It is, however, widely 
distributed, and has been obtained in Dorset, at Burton Bradstock, Broadwindsor, 
Clifton Maybank, Bradford Abbas, Halfway House and Combe Quarry near Sher- 
bourne. In Somerset, at Misterton, Haselbury and at Crewkerne Station ". The 
B.M.(N.H.) material, on which this revision is based, confirms this distribution. 
The species has not been recorded in the literature outside Britain and the only 
possible occurrence known to the author is represented by a few poorly preserved 
specimens collected by him from the parkinsoni zone of Normandy. 

Occurrence. No details were given by Buckman and Walker, but as the species 
seems to be restricted to the Inferior Oolite, it presumably lived in a fairly high- 
energy environment. The material in the collection of the B.M.(N.H.) shows quite 



68 JURASSIC RHYNCHONELLIDS 

a wide range in the size of the individuals and as they are very well preserved this 
would indicate that they had not been carried far, if at all, before being buried. 

A. (Acanthorhynchia) spinulosa (Oppel) 
(PI. 6, figs. 8-9, text-fig. 25 (top)) 

1850 Hemithiris senticosa (pars) Orbigny: 375. 

1857 Rhynchonella spinulosa (pars) Oppel: 608. 

1904 Acanthothyris spinulosa (Oppel); Loriol: 284-85, pi. 27, fig. 44. 

1917 Acanthothyris spinulosa (Oppel); Rollier: 79. 

Emended diagnosis. Medium sized Acanthorhynchia ; transversely oval in out- 
line ; anterior commissure rectimarginate or with very slight uniplication ; beak 
small and incurved ; beak ridges and interarea present. 

Stratigraphical range. Oxfordian — exact horizons not known. 

Type specimen. A type specimen is not selected as this would have to be chosen 
from Orbigny's specimens of Hemithiris spinulosa (discussed below) and these are 
neither stratigraphically nor geographically accurately located. 

Material. 2 specimens, collection of University of Dijon from Mt. Saon (Haute 
Marne) France ; 6 specimens, collection of B.M.(N.H.) nos. B. 70044-49, from 
Haute Marne ; 2 specimens Orbigny collection. 

Description. External characters. The shell is equally biconvex and the width 
is greater than the length. The thin test is evenly covered with fine bifurcating ribs 
bearing numerous spine bases along their crests. The ribs are coarser and the spine 
bases more widely spaced than in A. (Acanthorhynchia) panacanthina. 

Beak ridges are developed and limit the small but distinct interarea ; the small, 
incurved beak is adpressed to the brachial valve and obscures the delthyrium. 
Deltidial plates are disjunct and only weakly developed ; they appear briefly in a 
few of the anterior serial sections. Weakly developed growth lamellae, but not 
growth lines, may be present. 

Dimensions of Figured Specimens. 

length thickness width 

2-30 cm 1-40 cm 2 54 cm 

205 cm i-io cm 2-35 cm 

Internal characters. Pedicle valve. Dental lamellae are completely lacking and 
the teeth are inserted laterally at a rather shallow angle. 

Brachial valve. The most distinctive internal feature is the presence of what 
appears to be a cardinal process, developed from the inner hinge plates. The long, 
straight sockets, as seen in section 1-9 of text-fig. 25 appear to be distinctive. The 
crura are radulifer and curve slightly towards the pedicle valve. The median septum 
is poorly developed and is present only as a median ridge for most of its length. 

Distribution. The only records are from the Haute Marne area of France and 
from the Ledonien Jura. 

Remarks. The name spinulosa was first used by Oppel (1857) when he cited it 
as a " new name " for Hemithiris senticosa Orbigny in a list of Oxfordian fossils ; 
unfortunately, he did not figure or state exactly what he meant by it. 



FROM NORTHWESTERN EUROPE 



69 



Rollier (1917) suggested that the name spinulosa had been used by previous 
authors for two quite distinct species ; one of which he considered to be " Argovien " 




0-3 ^?M 



0-7 



/0-9 



' ^ \m 



10 



'1-2 




--'-^-.''1-6 



^ & 



/ 

2 6 



2 2 



_,-' 1-9 



1-7 



A 



30 



3-6 



39 



nQQQQi 




29 "--■ — -— -" 31 

Fig. 25. top. Transverse serial sections of A. (Acanthorhynchia) spinulosa (Oppel). Oxfordian. 
Mont Saon, (Haute Marne), France. Specimen donated by the University of Dijon. ( X4). 
bottom. Transverse serial sections of A. (Acanthorhynchia) sp. Kimmeridgian. Oignon 
Limestone, (Ager & Evamy, 1963), Colomieu (Ain), France. (X4). 



70 JURASSIC RHYNCHONELLIDS 

and the other " Oxfordien " in age. To remedy this he proposed that the name 
spinulosa be restricted to the larger, oval species from the " Argovien " and that the 
smaller, pyriform species of the " Oxfordien moyen et sup6rieur " be designated 
Acanthothyris lorioli sp. no v. Rollier's reason for preferring to restrict the name 
spinulosa to the larger form was that he considered Oppel's original designation 
inadequate and that, of the two species in question, this seemed to him to be the 
more likely to have been known to Orbigny, judging from the localities cited in the 
" Prodrome ". 

Plaster casts of the specimens labelled Hemithiris senticosa in the Orbigny collec- 
tion at the Jardin des Plantes, Paris, were kindly made available to the author by 
the curator, Dr. Fischer. These specimens are in two boxes labelled 3782, contain- 
ing three specimens from Chatel Censoir (Yonne) and 3782a, containing two speci- 
mens from Nantua (Ain). Both of the collection numbers contain a specimen of 
both A. (Echinirhynchia) lorioli and A. (Acanthorhynchia) spinulosa sensu Rollier. 
None of the specimens is referable to A. senticosa (Schlotheim). One of the speci- 
mens labelled 3782 probably belongs to a third, undescribed species and closely 
resembles a form collected by Dr. Evamy from the southern French Jura of Ain. 
Unfortunately not enough specimens of this form have been collected to make a full 
taxonomic investigation possible. 

From the investigation of the original Orbigny material, described above, it seems 
that neither Oppel nor Rollier realised the diversity of specimens included by 
Orbigny under the name Hemithiris senticosa. However, the author would agree 
with Rollier's conclusions and these are here followed. 



A. (Acanthorhynchia) vilsensis sp. n. 

(PL 6, figs. 10-12, text-fig. 26 (bottom)) 

1886 Rhynchonella myriacantha Deslongchamps; Rothpletz: 156, pi. 12, figs. 24-27. 

Name. From the type area of the " Vilser Alp " in the Tirol. 

Diagnosis. Medium sized Acanthorhynchia with rather coarse ribbing ; beak 
suberect ; uniplication in anterior commissure. 

Stratigraphical range. Given by Rothpletz as Callovian — macrocephalus to 
lamberti zones. 

Type specimen. The holotype designated is the specimen figured by Rothpletz 
(1886) pi. 12, fig. 25, which came from Legam bei Vils in the Tirol ; this specimen is 
now in the Bayerische Staatsammulung fiir Palaontologie und historische Geologie 
in Munich, and is numbered AS xiv 2. 

Material. 17 specimens from the Bayerische Staatsammulung. 

Description. External characters. The adult specimens are biconvex, trans- 
versely oval in outline and with the width always greater than the length, while the 
juvenile forms have a subcircular outline. The beak is suberect ; beak ridges and 
interarea are only very weakly developed. In nearly all the specimens examined, 
the beak had been broken off about level with the top of the brachial valve. In a 
few more complete specimens, the delthyrium appears to be open but this feature 



FROM NORTHWESTERN EUROPE 71 

has not been satisfactorily observed. The test is evenly covered with bifurcating, 
spine-bearing ribs which are quite coarse, as compared with other species in the 
genus. A few growth lamellae are frequently present but growth lines have only 
been observed on one specimen. The anterior commissure has a characteristic low 
uniplication. 

Dimensions of Figured Specimens. 

length thickness width 

170 cm 1 -oo cm 2-00 cm 

1-53 cm 0-93 cm 1-67 cm 

1-33 cm 070 cm 1-43 cm 

Internal characters. The specimen sectioned, which appeared to be of typical 
preservation for the locality, was very badly recrystallized and with the shell almost 
completely missing. The effect of this has been to obliterate the fine structure 
together with the crura which were rapidly lost in a mosaic of crystalline calcite ; 
the predominant lack of fine external ornament may also be a result of the poor 
preservation. 

Pedicle valve. Slender, slightly diverging dental lamellae are present and bound 
a subquadrate delthyrial cavity. The teeth are large with distinctly flattened bases. 

Brachial valve. The socket floors are very unusual in that they come to lie in 
a position sub-parallel to the plane of articulation. Both inner and outer socket 
ridges are well developed. 

Distribution. The species is only recorded from the type locality. 

Occurrence. In his original remarks, Rothpletz described it as occurring fairly 
abundantly in a " Brachiopoden-Lumachelle " together with numerous molluscs. 

Remarks. The species differs from A. myriacantha, as figured by Deslongchamps 
(1859), m being larger, having a uniplication in the anterior commissure and in having 
much coarser ribs on which the spine bases are more sparsely distributed. The small 
disjunct deltidial plates figured by Deslongchamps have not been observed in A. 
(Acanthorhynchia) vilsensis. 



A. (Acanthorhynchia) sp. 

(PI. 7, figs. 10-11, text-fig. 25 (bottom)) 

Material. 4 specimens including B.M. BB. 44168, BB. 44169. 

Remarks. The material was collected from the thin bedded Oignon Limestone* 
(Ager and Evamy, 1963) near Colomieu (Ain). Both externally and internally, the 
species shows considerable resemblance to A. (Acanthorhynchia) spinulosa ; how- 
ever, the difference in stratigraphical horizon and the variability shown by the 
limited material available suggests that the species should be regarded as being 
derived from the former rather than conspecific with it. The largest specimen 
collected resembles one of the forms in the Orbigny Collection labelled Hemithiris 
senticosa. 

* Oignon Limestone is of Upper Kimmeridgian age. 



72 



JURASSIC RHYNCHONELLIDS 





/ 



/ 



N 



\ 



\ 



\ 






\ V 



/ \ 



19 



\ 



9 



? 



/ 



2-1 



Fig. 26. top. Transverse serial sections of A . (Echinirhynchia) senticosa (Schlotheim). " Malm 
delta ". Heiligenstadt, Germany. Specimen donated by the University of Erlangen. ( X 8). 
bottom. Transverse serial sections of A. {Acanthorhynchia) vilsensis sp. nov. Topotype. 
Callovian. Legam bei Vils in the Tirol. Specimen donated by the Bayerische Staatsamm- 
lung. (x8). 



FROM NORTHWESTERN EUROPE 73 

ECHINIRHYNCHIA subgen. nov. 

Name. Latin echinus-im, the sea urchin : the shell of Echinirhynchia is covered 
with spines. 

Type species. Terebratnlites senticosus Schlotheim (1820). 

Diagnosis. Small to medium sized species of Acanthorhynchia ; usually sub- 
triangular in outline ; valves equally biconvex or with a flattened brachial and more 
inflated pedicle valve ; beak small and either adpressed to the brachial valve or 
suberect ; deltidial plates reduced or absent ; when the beak is suberect there is an 
open triangular delthyrium ; no dental lamellae or septalium ; anterior commissure 
rectimarginate. 

Stratigraphical range. Oxfordian to Lower Volgian : mariae — Gravesia 
zones. 

Distribution. The subgenus appears, from the literature, to be restricted to the 
Jura, ranging from the southern French Jura through Switzerland to the Franconian 
Alb north of Erlangen ; the only records outside this area are of A . [Echinirhynchia) 
fileyensis (Buckman & Walker) from the Lower Calcareous Grit of Filey Brigg, 
Yorkshire, and of " Acanthorhynchia aff. senticosa (Schlotheim) " from Rumania 
(Patrulius, 1964). 

Remarks. The three species included are the only ones of which sufficient 
material has been obtained to allow description. Several nominal species, such as 
Terebratula senticosa impressae Quenstedt and T. senticosa y alba Quenstedt, 
almost certainly belong within the subgenus but the author has not been able to 
evaluate their position. The Quenstedt species mentioned above are the only ones 
recorded from the Swabian Alb area, although some of Oppel's records of Rhyn- 
chonella spinulosa probably included material referable to Echinirhynchia. 



A. (Echinirhynchia) senticosa (Schlotheim) 
(PI. 7, figs. 1-3, text-fig. 26 (top)) 

1820 Terebratnlites senticosus Schlotheim: 268. 

1834 Terebratula senticosa Buch: 59-60. 

1838 Terebratula senticosa Buch: 162-63, pi. 16, fig. 5. 

1858 Terebratula senticosa Quenstedt: 457. 

1871 Terebratula senticosa silicea Quenstedt: 115, pi. 39, fig. 66. 

1886 Rhynchonella senticosa (Schlotheim); Rothpletz: 93, 156-57, pi. 15, figs. 4-9. 

1889 Acanthothyris senticosa (Schlotheim); Buckman and Walker, 55-56. 

1917 Acanthothyris senticosa (Schlotheim); Rollier: 80. 

1918 Acanthorhynchia senticosa (Buch); Buckman, 70. 

Emended diagnosis. Small to medium sized, pyriform Echinirhynchia ; test 
covered with numerous fine, spinose ribs ; beak relatively large and suberect to 
erect ; deltidial plates absent resulting in an open triangular delthyrium ; recti- 
marginate. 

Stratigraphical range. Kimmeridgian — Lower Volgian : pseudomutabilis to 
Gravesia zones. 



74 J URASSIC RHYNCHONELLIDS 

Type specimen. Neotype (pi. 7, fig. 1), here designated, is the specimen figured 
by Rothpletz (1886) pi. 15, fig. 5, which is No. AS vii 330 in the collection of the 
Bayerische Staatsammlung fur Palaontologie und historische Geologie in Munich. 
It is considered desirable to designate a neotype as the name senticosa has been used 
very loosely in the past for widely differing forms. 

Material. 12 specimens from the collection of the University of Erlangen ; n 
specimens from the Bayerische Staatsammlung, Munich. 

Description. External characters. The following is a translation of the original 
description by Schlotheim, " Partly in very complete examples, with perfectly 
preserved shells, from Grumbach, near Amberg, in the Pfalz, petrified in chert and 
probably belonging to the Jura formation. This very rare and remarkable Tere- 
bratulite, only recently discovered, has a somewhat elliptical, almost pear-shaped 
form, it is not particularly thick, is regularly convex with a fine, evenly striated, 
shagreen like surface and on both valves the somewhat prominent rays are covered 
with numerous, small sharp spines, by which this and spinosus may be distinguished 
from all other known species. It is very seldom found ". 

The Schlotheim description is quite accurate and mentions the two most distinctive 
characters, namely the pyriform shape and the fine, spine-covered ribs. The ribs 
characteristically bifurcate laterally thereby maintaining an even distribution over 
the surface. The species is also characterized by the possession of a rectimarginate 
anterior commisure. The beak is suberect and does not obscure the triangular 
delthyrial opening ; no deltidial plates have been observed. Beak ridges and inter- 
area are not developed. Some specimens have a few growth lamellae developed, but 
growth lines are not present ; it is possible that the latter may have been obliterated 
as a result of the silicification suffered by all the specimens examined. The valves 
are about equally biconvex, although the brachial valve tends to be rather flat and 
may be the less inflated. 

Dimensions of Figured Specimens. 

length thickness width 

1-35 cm 061 cm 1-17 cm 

1-94 cm 087 cm 1 62 cm 

1-35 cm 067 cm 1-22 cm 

The last set of figures are the dimensions of the neotype. 

Internal characters. The specimen sectioned was from Heiligenstadt, Franconia, 
and had a silicified shell and dolomitised infilling. This preservation has resulted in 
the loss of all fine detail and it has only been possible to give rather generalized 
drawings of the internal structures. 

Pedicle valve. There are no dental lamellae and the large teeth are inserted 
laterally at a shallow angle. 

Brachial valve. A cardinal process is present. Inner and outer socket ridges 
are both well developed. The form of the crura could not be determined as they 
disappeared into the recrystallized matrix immediately below the last section shown 
in text-fig. 26 — top. 

Distribution. The Franconian and Swabian Jura. 



FROM NORTHWESTERN EUROPE 



Occurrence Dr. Zeiss, of Erlangen, has stated (personal communication io6«| 
that the material from Heiligenstadt occurs in the " Sohwammfacies' There i 

ReZe 3 ,0n T a h Vailable C ° nCeming thC mate ™' fl ^ red * Ro'Me (.886) " 

been Zcanse of th P eT" f aS rem E ked ^"^n, is rare and this has probably 

aonhed to H„ *l ," ^° at Wh ' Ch Spedes the name *"*«»« should be 

original lo C alit?d.ed e bv ^ViT^ ^^^ figUfed * Roth P ktz <™» tta 

thl, iT y _, y Schlothelm seem to fit the latter's description so exactlv 

hat there seems no doubt that this was the species intended by him As can be Vl„ 

d Z * H lgmal t SC u ipti ° n ' "° eMCt ««4al horizon waJgrTn and some of ne 
difficulty has undoubtedly arisen through Buch d8,8> ritW ?£ sTi ?e • 
locality for bis TereiratuU senUcosa and sfating fhe' rfon or^th s oealfiy toT 

Unn . r T f CTal aUthorS t0 reect the nal ™ «««««• when describing 

i4. (Echinirhynchia) lorioli (Rollier) 
(PL 7, figs. 7-9, text-fig. 27) 

1850 Hemithiris senUcosa (pars) Orbigny: 375 

1857 Rhynchonella spinulosa (pars) Oppel) : 608 

1871 ?Terebratula y alba Quenstedt: 114, pi. 39 , w 6 ,_ 6 , 

1897 Acanthothyvts spinulosa (Oppel); Loriol: 146.pl i 7 fi g I7 

1904 Acanthothyris spinulosa (Oppel); Loriol: 284-85' pi 27 fiJ*., 

1917 Acanthothyris Lorioli Rollier: 79. P 7 ' g ' 45 ' 

191 7 ? Acanthothyris dealbata Rollier: 80. 

Bic E o™ E ^ r"' P r f ° rm ° r SUb " CirCular ^mrkynchta. 

oiconvex , rectimarginate ; beak adpressed to the brachial valve 

^«To A n P e HICAL RANGE - ° Xf ° rdlan - the °*y exact records are from the trans- 

LorionTSoT'T^' fi LeCt0t yP e ' here select ^. the original specimen figured b V 
^ Pl : I7 ' ^ ^ collected from Montfaucon (Jura bernois). g Y 

RR Ifrt V 7 !P eC1 T menS fr ° m the collec tion of the B.M.(N.H.) two BB 44166 

Frenl 7' *' JUra ; Z ° SpedmenS Collected b y Dr - Enay from thf central 

French Jura ; 2 specimens collected by Dr. Evamy from Ain 

specSensTow ?%£? ??"*?•*', ^ SpedeS iS ^^ blconvex bu t some 
specimens show a flattened brachial valve. The test is covered with many fine 

bifurcating ribs bearing numerous fine spines. The beak is small and closely pressed 

tt^^^ 01 ^ 11 ^ thC delthynUm ; bGak ridgGS and -terar P ea S not 
oble ved C are SOmetimes P rese "t but growth lines have not been 



76 jurassic khynchonellids 

Dimensions of Figured Specimen. 



length 
1-54 cm 



thickness 

075 cm 



width 
i 40 cm 



04 





<& <& 






1-2 




3 2 



FlG. 27. 'I"ransverse serial sections of A. (Echinirhynchia) lorioli (Kollier). Oxfordian- 
transversarium zone. St. Sorlin, Liefnans (Jura), France. (x8). 



FROM NORTHWESTERN EUROPE 77 

Internal characters. Pedicle valve. There are no dental lamellae. The deltidial 
plates are disjunct and only weakly developed. The teeth are strong and inserted 
laterally at a rather shallow angle. 

Brachial valve. A cardinal process is not present but the species exhibits the 
thickened inner hinge plates characteristic of the genus. The crura were not well 
preserved but are possibly of the calcarifer type. A median septum is only present 
as a low ridge. 

Distribution. The species has only been definitely recorded from the central 
and southern French Jura. 

Occurrence. The material collected by Dr. Enay was obtained from dark- 
coloured clays where it was associated with Monticlarella triloboides. 

Remarks. The specimen sectioned came from St. Sorlin (Jura) one of the locali- 
ties mentioned by Rollier in his original description. 

The separation of A. (Echinirhynchia) lorioli from A. (Acanthorhynchia) spinulosa 
is described under the latter. 

A. (Echinirhynchia) fileyensis (Buckman & Walker) 
(PL 7, figs. 4-6) 

1889 Acanthothyris senticosa var '. fileyensis Buckman and Walker: p. 56. 

Emended diagnosis. Small to medium sized Echinirhynchia ; beak small and 
suberect to erect ; valves equally biconvex. 

Stratigraphical range. Oxfordian — mariae and lower cor datum zones. 

Type specimen. Lectotype, here selected, specimen no. B. 31867 in the Walker 
Collection of the B.M.(N.H.), collected from the Lower Calcareous Grit of Filey, 
Yorkshire. 

Material. 6 specimens from the Walker Collection in the B.M.(N.H.) ; nos. 
B. 31313, B. 31335, B. 31275 ; 2 specimens from the collection of the Naturhis- 
torisches Museum, Basel. 

Description. External characters. The shell is biconvex and pyriform in out- 
line with a straight anterior margin. The ribs are fine but well marked and covered 
with spine bases. The beak is small and erect or sub-erect ; beak ridges and inter- 
area are not developed. Details of deltidial plates, if any, and delthyrium not 
observed. 

Dimensions of Figured Specimen. The following measurements are those of 
the lectotype. 

length thickness width 

080 cm 0-37 cm 0-74 cm 

Internal characters. Not investigated through lack of material. 

Distribution and occurrence. The species has only been definitely recorded 
from Filey, Yorkshire, where it occurs in the Lower Calcareous Grit. The only other 
possible conspecific material known to the author is two specimens from the 
" Renggeri-Thon " of Baden, in the collection of the Naturhistorisches Museum, 
Basel and numbered L 1701/1-2. 



78 JURASSIC RHYNCHONELLIDS 

Remarks. The form was first mentioned by Buckman and Walker in 1889 in 
their discussion of the spinose Rhynchonellidae, when they described it under the 
name of Acanthothyris senticosa var. fdeyensis and suggested that it resembled A. 
senticosa of Orbigny. However, as A. (Echinirhynchia) fdeyensis shows marked 
differences from both the specimens in the Orbigny Collection and A. {Echinir- 
hynchia) senticosa (Schlotheim), it has been decided to describe it as separate species 
despite the paucity of material. 

Subfamily TETRARHYNCHIINAE Ager, 19656 
Genus SOMALIRHYNCHIA Weir 

1925 Somalirhynchia Weir: 79. 
1929 Somalirhynchia Weir; Weir: 38-39. 
J935 Somalirhynchia Weir; Muir-Wood: 93. 
1964 Praecyclothyris Makridin, 150-51. 
19656 Somalirhynchia Weir; Ager: H614. 

Type species. S. africana Weir, by original designation. 

Emended diagnosis. Large, subpentagonal, trilobate rhynchonellids ; 20-30 
coarse, simple, subangular ribs ; large, sub-erect beak ; possesses septalial plates 
and median septum, which together form a septalium ; crura radulifer. 

Stratigraphical range. Upper Oxfordian — Lower Kimmeridgian. 

Distribution. Somaliland, Syria, N.W. Europe, Russia. 

Remarks. Weir (1925, 1929) gave the name S. africana to the species figured by 
Noetling (1886) as Rhynchonella moravica Uhlig ; he suggested that the differences in 
outline and particularly in muscle scar patterns necessitated regarding R. moravica 
Noetling as differing from R. moravica Uhlig not only specifically but generically. 
The author considers that Weir was probably correct in establishing a new species 
for Noetling's R. moravica. However, in internal structure, as well as general 
external appearance, S. africana, in the author's opinion, closely resembles R. 
moravica Uhlig and should be regarded as congeneric with it. 

In basing his taxonomy almost entirely on the muscle scar patterns, Weir was 
closely following the ideas of Buckman. However, it is now considered that the 
similarity exhibited by all the other features outweigh in taxonomic significance the 
dissimilarity of the muscle patterns. The internal details of R. moravica Uhlig were 
figured by Wisniewska (1932) and, allowing for the differences in appearance caused 
by her angle of sectioning, are closely comparable to those given by Muir-Wood 
(1935) for S. africana. They appear to have little in common with Septaliphoria 
arduennensis (Oppel), the type species of Septaliphoria Leidhold, to which genus R. 
moravica Uhlig was assigned by Wisniewska. Consequently, the author does not 
understand on what grounds Muir-Wood (1935) remarked, " Further study of the 
internal characters of R. moravica by Wisniewska (1932, p. 22) has emphasized the 
differences between the two species (i.e. R. moravica Uhlig and 5. africana Weir). 
R. moravica is now assigned to the genus Septaliphoria (Leidhold 1920) ". 

The wide geographical distribution of the genus is coupled with a very limited 
stratigraphical range. It is considered that the distribution of the genus can be 



FROM NORTHWESTERN EUROPE 79 

extended to Russia, as it appears from Makridin's figures that some at least of his 
species of the genus Praecyclothyris should be attributed to Somalirhynchia. That 
the genus has not previously been recorded from Europe is probably the result of its 
name, which has possibly been taken by previous authors as implying a restricted 
geographical occurrence. 



Somalirhynchia sutherlandi (Davidson) 
(PI. 7, figs. 12-16, text-fig. 29) 

1873 Rhynchonella Sutherlandi Davidson: 196, pi. 8, figs. 1-2. 

1878 Rhynchonella Sutherlandi Davidson; Davidson: 190-91, pi. 25, figs. 5-8. 

1917 Rhynchonella Sutherlandiae Davidson; Rollier: 172. 

1 91 8 Rhynchonella Sutherlandi Davidson; Buckman: 51. 

Emended diagnosis. Medium to large sized Somalirhynchia ; 15-25 coarse, 
simple, angular ribs ; usually trilobate ; suberect beak, septalium present ; crura 
radulifer. 

Stratigraphical range. Kimmeridgian — probably restricted to the upper 
mutabilis and pseudomutabilis zones. 

Type specimen. Lectotype, here selected, the specimen figured by Davidson 

(1878), pi. 25, fig. 5)- 

Material. 19 specimens from the collection of the Geological Survey of Scotland, 
together with the very limited material available at the B.M.(N.H.) nos. B. 29749, 
B. 29752, B. 22586, B. 29751, B. 29753. A recent attempt by the author to collect 
further material yielded a single fragmentary specimen. 

Description. External characters. The species is distinctive on account of the 
very large size attained by many individuals ; Davidson (1878) claimed it to be the 
largest Mesozoic rhynchonellid. The outline is subpentagonal and the shell is 
biconvex with the brachial valve the more inflated. It is generally trilobate but a 
few specimens have been observed displaying quite marked asymmetry. Width, 
according to Davidson, is always greater than length, and while this seems to be 
correct the crushed nature of much of the material makes this difficult to ascertain 
in more than a few cases. Plate 7 figures the best preserved material available. 

The strong, suberect beak is flanked by well defined beak ridges which limit the 
incurved interarea. A relatively small pedicle opening is present ; the deltidial 
plates, as seen in transverse section, are strongly conjunct. Fine, concentric growth 
lines have been observed on one specimen ; this may well reflect the poor preserva- 
tion rather than the genuine lack of such ornament on most specimens. 

Dimensions of Figured Specimen. 

length thickness width 

2-67 cm 2-33 cm 3-16 cm 

Internal characters. The specimen sectioned was slightly crushed, resulting in 
do rso- ventral flattening of the shell. The outline of some of the internal features 
was partly distorted by pressure solution caused by the coarse, quartz matrix. 



8o 



JURASSIC .RHYNCHONELLIDS 

Pedicle valve. A pedicle collar is present ; the deltidial plates are conjunct and 
" crumple up " against one another. A quadrate delthyrial cavity is flanked by 



OCSCUD 





1 " JL "' 

^~~4 mk"-* ^^ /~^ 

JL " 1" 



X«X 



1 01 



|' S 2 



0.1 



oT5 



Fig. 28. Transverse serial sections of Somalirhynchia africana Weir. Taken from Muir-Wood 

(1035. %• 8, p. 95)- 



FROM NORTHWESTERN EUROPE 81 

long, stout, subparallel dental lamellae. The hinge apparatus is strongly developed 
with massive crenulated teeth and strong lateral denticulae. 

Brachial valve. The most obvious feature is the presence of a median septum 
and well developed septalial plates forming a septalium. The crural bases do not 
become apparent, as seen in transverse section, until after the hinge plates have 
divided and the septalial plates disappeared. The crura are radulifer. The sockets 
are crenulated and both the inner and outer socket ridges are strongly developed. 

Distribution. The species has only been recorded from the Helmsdale area of 
Sutherland, Scotland. 

Occurrence. The only known occurrence is in the Kimmeridgian " Boulder 
Beds " of the Helmsdale area, east Sutherland, Scotland. Bailey and Weir (1935) 
suggested that the rhynchonellids, together with the terebratulids and corals, owed 
their presence in these beds to the effects of submarine faulting. They suggested a 
near-shore fault scarp which, " separated a comparatively shallow- water facies, 
characterized by rounded pebbles, sand, Rhynchonella, Terebratula, Ostrea, sea 
urchins corals, etc. from a comparatively deepwater facies, characterized by mud, 
debris of land plants, ammonites etc.". The rhynchonellids etc. were then swept 
over the scarp edge by the movements of the fault and the accompanying 
tsunamis. 

Ager (1965a), accepting this hypothesis, suggested that the S. sutherlandi 
inhabited, " sublittoral, non-depositional sea floors "; the rarity of the species then 
being accounted for by the infrequency with which such deposits are preserved. 
However, although no other representative of the genus occurs in Britain, S. moeschi 
(Haas) is present in the " Rauracian " of the French and Swiss Jura. Material of 
the latter species has been given to the author by Dr. Enay of the University of Lyon 
and he has stated (personal communication, 1965), that S. moeschi, at least within the 
Jura meridionale, always occurs in association with corals. From this it would also 
seem possible that S. sutherlandi inhabited shallow marine environments and was 
not necessarily restricted to rocky shores. The presence of corals at Helmsdale 
indicates the possibility of a reef environment being available. 

Remarks. It appears that S. sutherlandi has not been further described since 
Davidson in 1878 and during this time there have been only two tentative generic 
assignations. In the first, Buckman (1918), placed it in his genus Rhactorhynchia 
but although he indicated his uncertainty he did not discuss the matter. Ager 
(1965a), suggested the possibility of sutherlandi being in the genus Russirhynchia 
Buckman, which is otherwise restricted to the Russian Platform. The present 
author, however, prefers to attribute the species to the genus Somalirhynchia Weir. 
Reproduced below, text-figure 28, are the serial sections given by Muir-Wood (1935) 
for the type species, S. africana Weir, and it can be seen that they show a striking 
resemblance to those of 5. sutherlandi. Externally the species are also comparable 
in general shape, ribbing, and form of the beak and the pedicle opening. Both 
internally and externally " R." sutherlandi appears to be much closer to 5. africana 
than to the type species of Russirhynchia namely, R. fischeri (Rouillier). 



82 



JURASSIC RHYNCHONELLIDS 





5-3 



6 6 




5-9 



t 



V ^ 



7 4 



8 2 



r <\ 



8-5 



Fig. 29. Transverse serial sections of Somalirhynchia sutherlandi (Davidson). Topotype. 
Kimmeridgian. Boulder Beds. West Garty, Sutherland. Specimen donated by the 
Geological Survey, Edinburgh. ( X3). 



FROM NORTHWESTERN EUROPE 83 

Somalirhynchia moeschi (Haas) 
(PI. 8, figs. 1-3, text-fig. 30) 

1890 Rhynchonella trilobata Zieten var. Moschi Haas: 58-59, pi. 7, fig. 7, pi. 8, figs. 1-2. 

1893 Rhynchonella trilobata Zieten var. Moeschi Greppin: 98, pi. 7, figs. 3-4. 

1917 Rhynchonella Moeschi Rollier: 174. 

1932 ? Septaliphoria moravica (Uhlig) ; Wisniewska: 22-24, P^ 5> n S s - 3 _ 4- 

1964 Praecyclothyris moeschi (Rollier); Makridin: 158-60, pi. 6, fig. 8, pi. 7, fig. 1. 

Emended diagnosis. Medium to large, subpentagonal Somalirhynchia ; mark- 
edly trilobate with about 6-8 ribs on the fold ; septalium present although septalial 
plates relatively weakly developed ; crura radulifer. 

Stratigraphical range. Upper Oxfordian ; records, such as those of Haas 
(1890), from the Kimmeridgian are almost certainly the result of misidentifying 
Lacunosella trilobataeformis Wisniewska. The latter species strongly resembles 5. 
moeschi externally but can be readily differentiated by its possession of bifurcating 
ribs. 

Type specimen. Lectotype, here selected, the specimen figured by Haas (1890) 
pi. 7, fig. 1 ; the specimen was obtained from the " Wangener Schichten " of 
" Engelberg bei Olten ". 

Material. 80 specimens collected by Dr. Enay from the " departements " of 
Jura and Ain. 6 specimens from the collection of the University of Dijon labelled, 
" Corallien inferieur? ",7 specimens collected by the author from Pontarlier, near 
Besancon. 

Description. External characters. The species has a subpentagonal outline and 
is markedly trilobate. There are between 20 and 30 coarse, simple subangular ribs 
of which 6 to 8 lie on the fold. 

The sharp, suberect beak is flanked by weak beak ridges and these in turn delimit 
the slightly incurved interarea. The deltidial plates, which tend to project slightly 
around the circular, submesothyridid pedicle opening, are conjunct ; however, in 
many specimens the deltidial plates are missing, presumably through having fallen 
out after death, leaving an open delthyrium. 

Dimensions of Figured Specimens. 

length thickness width 

3-57 cm 270 cm 3-82 cm 

3-12 cm 2-18 cm 3-00 cm 

Internal characters. Pedicle valve. The lateral cavities are small and limited by 
slender, subparallel, persistent dental lamellae. The delthyrial cavity is quadrate. 
The teeth are crenulated ; lateral denticulae are only weakly developed. 

Brachial valve. A septalium is present ; although the median septum is quite 
strong and persistent, the septalial plates are only weakly developed. The radulifer 
crura have clearly differentiated crural bases. The hinge plates are arched ventrally. 

Distribution. The French and Swiss Jura ; ? Poland. 

Occurrence. The specimens collected by the author came from a limestone 
with a typical reef fauna, namely lamellibranchs, terebratulids, gasteropods and 



JURASSIC RHYNCHONELLIDS 




4 2 



4 6 



V 



*- "\ / 



/ 

6 6 



I r ^ 



\ 



/ 



7-1 



\ 



A 1 



\ 



5 4 




6 



/ 



/ 



7-5 



Fig. 30. Transverse serial sections of Somalirhynchia moeschi (Haas). Oxfordian- 
transversarium zone. Arinthod (Jura). (X3). 



FROM NORTHWESTERN EUROPE 85 

echinoids together with numerous branching corals. S. moeschi was the most 
common brachiopod, followed in numbers by the smooth terebratulids, while the 
ribbed terebratulids were the rarest element of the fauna. 

Remarks. In discussing Septaliphoria moravica (Uhlig), Wisniewska suggested 
that Rhynchonella trilobata var. Moschi was probably a synonym. Unfortunately, 
the author has not seen the original specimens of Uhlig, Haas or Wisniewska. 
Judging from the published figures and assuming these to be typical, it would appear 
that the specimens described by Wisniewska are conspecific with S. moeschi (Haas) 
and that " Rhynchonella " moravica Uhlig is probably a separate species, the latter 
being differentiated by its more trilobate form. However, " R." moravica, which 
occurs in the bimammatum zone is almost certainly congeneric. 

In 1964, Makridin divided R. moeschi Rollier into two subspecies, namely moeschi 
moeschi and moeschi donetziana. He then selected the latter as the type of his new 
genus Praecyclothyris ; however, as under the international rules of nomenclature 
the type must be moeschi moeschi ; it is suggested that Praecyclothyris is a synonym 
of Somalirhynchia Weir. As Somalirhynchia is not mentioned in the Russian 
" Treatise ", it is assumed that the rather obscure publications in which it has been 
described are not known to palaeontologists in the Soviet Union. 



Subfamily CYCLOTHYRIDINAE Makridin, 1955 
Genus SEPTALIPHORIA Leidhold 

1920 Septaliphoria Leidhold: 354-55. 

1932 Septaliphoria Leidhold; Wisniewska: 18-20. 

i960 Septaliphoria Leidhold; Makridin: 251-52. 

1964 Septaliphoria Leidhold; Makridin: 96-7. 

19656 Septaliphoria Leidhold; Ager: H619. 

Type species. Rhynchonella arduennensis Oppel, by original designation. 

Emended diagnosis. Medium sized, subpentagonal rhynchonellids ; 10-25 
coarse, simple, subangular ribs ; septalium present ; crura radulifer. 

Stratigraphical range. Oxfordian — PKimmeridgian. 

Description. External characters. The genus consists of ordinary looking 
uniplicate rhynchonellids, having a tendency towards asymmetrical development of 
the anterior commissure. The complete, bilobed asymmetry characteristic of 
Torquirhynchia is very rarely found and in most cases the asymmetry of the anterior 
commissure has resulted from displacement of the uniplication. Septaliphoria is 
biconvex with the brachial valve the more inflated ; there is no posterior smooth 
area. 

The well developed, suberect beak is flanked by beak ridges which limit an incurved 
interarea. A large, hypothyridid or submesothyridid pedicle opening is present ; 
the deltidial plates are usually, but not invariably, conjunct. 

Internal characters. Pedicle valve. The dental lamellae are subparallel and 
relatively short. The teeth may be slightly crenulated ; lateral denticulae are 
variably developed. 



86 JURASSIC RHYNCHONELLIDS 

Brachial valve. The most important feature is the septalium ; in his original 
description Leidhold (1920) regarded this structure as the diagnostic feature of the 
genus. 

The crural bases are clearly differentiated and are distinctive in that they can be 
seen at the same level as the septalial plates in transverse section. In both the 
species sectioned, the crura ended with the curiously shaped, but consistent, pro- 
cesses seen in text-figs. 31-33. 

Species. The following nominal species are attributed to the genus : 

S. arduennensis (Oppel) (1858, pp. 615, 639, 654) 

?S. hudlestoni (Rollier) (1917, p. 172) 

S. paucicosta sp. nov. 

5. pinguis (Roemer) (pars in Wisniewska 1932, pp. 24-29, pi. I, figs. 1-18) 

IS. septentrionalis sp. n. 

?S. pectunculoides (Etallon) (in Makridin 1964, pp. 102-5, pi- I» fig s - 4~6) 

S. sobolevi Makridin (1964, pp. 97-99, pi. I, figs. 16-18). 
Distribution. The taxonomic confusion which has surrounded the genus, 
together with the lack of published serial sections, makes identification from the 
literature very difficult. However, the genus occurs in Yorkshire, northern 
France, the Swiss Jura and ? Dorset. Some of Makridin 's species of Septaliphoria 
from the Russian Platform also undoubtedly belong to the genus, as here defined. 

Remarks. Much confusion has arisen about the exact nature of the genus, 
mainly as a result of two factors. Firstly, Leidhold (1920) chose as his type species 
the poorly defined Rhynchonella arduennensis of Oppel, and secondly, because 
Leidhold's description of the septalium was based on a misunderstanding regarding 
its development. 

The synonymy of the type species is described under S. arduennensis. A full 
description, with figures and sections of the septalium has already been given in the 
section on morphology and it is only necessary to repeat here that it forms through 
the fusion of the septalial plates with the septum and not as a result of a bifurcation 
of the septum, as originally suggested by Leidhold. 



Septaliphoria arduennensis (Oppel) 
(PI. 8, figs. 4-5, text-figs. 31-32) 

1858 Rhynchonella Arduennensis Oppel: 615, 639, 654. 

1871 Terebratula inconstans Quenstedt: 141, pi. 40, fig. 57. 

1917 Rhynchonella Arduennensis Oppel; Rollier: 171-72. 

1920 Septaliphoria arduennensis (Oppel); Leidhold: 354, pi. 5, fig. 2. 

1932 Septaliphoria arduennensis (Oppel); Wisniewska: 18. 

Emended diagnosis. Medium sized subpentagonal to subtriangular Septali- 
phoria ; biconvex ; frequently slightly asymmetrical ; about 20 simple, subangular 
ribs ; well developed septalium crura radulifer. 

Stratigraphical range. Lower Oxfordian ; cordatum and lower transversarium 
zones. 



FROM NORTHWESTERN EUROPE 87 

Type specimen. A neotype is here proposed in view of the uncertainty which has 
surrounded this important species. Although the specimen chosen is not from any 
of the localities cited by Oppel (1858) it is from the horizon mentioned by him, 
namely the " Terrain a Chailles ", and has been selected to conform as nearly as 
possible to the specimen figured by Leidhold, as he was the first to figure the species 
as such and, consequently, his figure has been accepted as definitive both of the 
species and of the genus Septaliphoria of which it is the type. Unfortunately, it is 
not possible to collect material from the area around Mezieres mentioned by Leidhold, 
as the Oxfordian is no longer exposed there (Dr. Maubeurge, personal communication 
1965). As the specimen figured by Leidhold shows the internal structure, it seems 
probable that it was a silicified specimen from the " Terrain a Chailles ". 

The author has not been able to find the type specimens of either Oppel or Leidhold 
and it is considered highly unlikely that they still exist in a recognizable form. The 
neotype was collected from the " Terrain a Chailles " at Pagny-sur-Meuse, near 
Nancy, France. 

Dimensions of neotype. Length 2-33 cm ; thickness 1-90 cm and width 
2-60 cm. 

Material. 25 specimens from Pagny-sur-Meuse (Meurthe et Moselle), including 
BB.44173, BB.44174. 

Description. External characters. The outline of adult specimens is sub- 
pentagonal while that of the smaller ones is subtriangular ; the smaller specimens 
also tend to be equally biconvex while the larger ones have a more inflated brachial 
valve. 

The beak is suberect and flanked by fairly well developed beak ridges. The inter- 
area is relatively large and either flattened or slightly incurved. The deltidial plates 
are either disjunct or just conjunct. A large, well developed, circular pedicle 
opening is present and was presumably functional at all stages. Concentric orna- 
ment appears to be restricted to occasional, poorly developed growth lamellae. 

Dimensions of Figured Specimens. 

length thickness width 

2-49 cm 173 cm 2 67 cm 

2-33 cm 1-90 cm 2-60 cm 

Internal characters. All the specimens available were silicified and much of the 
fine detail of the internal structure has been lost. 

Pedicle valve. The narrow, elongated lateral cavities are delimited by slender, 
ventrally divergent dental lamellae which break away from the ventral wall of the 
valve well to the posterior of the plane of articulation. The teeth are strong and 
crenulated ; lateral denticulae are present. 

Brachial valve. A septalium is present and can be clearly seen in text-fig. 32. 
It is less distinct in the other specimen sectioned, probably as a result of partial 
silicification. The crural bases are distinct and can be seen lying immediately 
behind the septalial plates in section 1, 5 of text-fig. 31 ; the crura are of the radulifer 
type. The hinge apparatus is strong with crenulated sockets and both inner and 
outer socket ridges well developed. 



ss 



JURASSIC RHYNCHONELLIDS 



Distribution. The distribution of the species is very uncertain and the only 
definite records are from the " Terrain a Chailles ", of northern France and the Swiss 
Jura around Basel. 

Occurrence. The specimens collected by the author at Pagny were from sandy 
limestones containing cherty bands and nodules. The large associated fauna con- 
sisted of ammonites, lamellibranchs, mainly Pholadomya and Lima types, and 
Thurmannella obtrita. T. obtrita was the most abundant element and tended to 




3 8 



3 6 



Fig. 31. 



Transverse serial sections of Septaliophoria arduennensis (Oppel). Oxfordian. 
" Terrain a Chailles ". Pagny-sur-Meurse (Meurthe et Moselle). ( X3). 



FROM NORTHWESTERN EUROPE 



occur in groups while S. arduennensis occurred as single specimens and was much 
less common. 

Remarks. The name arduennensis first appeared in the literature in 1858 when 
Oppel used it as a " n. sp." for, " Rh. inconstans d'Orb. 1848, Prodr. 13 460 (non 
Sow.) ", in a list of Oxfordian fossils. He gave no description but subsequently 
included it in faunal lists from the ironstone of the Ardennes, as developed at Vieil- 
Saint Remy and Neuvizi, which he was describing together with other localities of 
his Ammonites biarmatus zone. However, he stated that he considered the above 
localities to be possibly of a higher horizon. Oppel's other citation of the species is 
from the " Zone des Cidaris florigemma. (Terr, a Chailles und unteres Coralrag) ", 
of the Swiss Jura. 




shell completely 

silicif ied 



6-5 



Fig. 32. Transverse serial sections of Septaliphoria arduennensis (oppel). Oxfordian. 
" Terrain & Chailles ". Pagny-sur-Meurse (Meurthe et Moselle), France. ( X 3). 



go JURASSIC RH YNCHONELLIDS 

Arkell (1956) thought the Neuvizy Ironstone to be of cor datum age, thus making it 
the lateral equivalent of the Terrain a Chailles which is cordatum and lower trans- 
versarium zones in age. This would then restrict arduennensis, as it was presumably 
conceived by Oppel, to those zones. 

Oppel proposed a new name for Orbigny's R. inconstans as he realized that the 
Oxfordian species mentioned by Orbigny was not the one which had already been 
described by Sowerby (1821) under the name inconstans from the English Kimmer- 
idgian. As Oppel was proposing arduennensis as a new name for R. inconstans of 
Orbigny, it seems reasonable to assume that the species to which he was referring 
exhibited some degree of asymmetry. Therefore, it seems fairly certain that by 
arduennensis he meant the medium sized asymmetrical rhynchonellid occurring quite 
frequently in the Lower Oxfordian of the areas mentioned by him. As can be seen 
from the figured specimens, the species shows varying degrees of asymmetry but is 
occasionally perfectly symmetrical. The latter condition is shown by the specimen 
figured by Leidhold and this again has lead to some confusion. The species does not 
appear to have been figured other than by Leidhold and as Terebratula inconstans 
(pars) by Quenstedt. 

In making arduennensis the type species of his new genus septaliphoria, Leidhold 
gave no description and merely commented that it came from the " Oxford-Stufe ". 
The caption to the one specimen figured described it as coming from the " Oxford- 
Stufe " of Mezieres. 

Rollier (1917) suggested the possibility of R. arduennensis Oppel being a synonym 
of Terebratulites helveticus of Schlotheim (18 13). The latter species was figured by 
Schlotheim but not given any description other than being listed as occurring in the 
" Jurakalkstein " near Basel and being given the citation " Scheuchzer (1718), fig. 
105 ". The latter figure is virtually indeterminable except as a ribbed brachiopod. 
Although Rollier maintained that R. helvetica " est incontestablement une forme 
asymmetrique de l'Oxfordian super, et du Rauracien inferieur, tres, frequente dans 
le Jura aux environs de Basle," this statement seems to be erring on the side of 
optimism in view of inadequacy of the figures and the lack of detail given by both 
Scheuchzer and Schlotheim. Quenstedt (1871) considered the Schlotheim species 
to be his Terebratula lacunosa multiplicata from the " Malm gamma "; this does not 
appear likely but at least demonstrates the uncertain nature of the species. It there- 
fore seems best to treat T. helveticus Schlotheim as a nomen dubium. 

The Terebratula helvetica Schlotheim figured by Zieten (1831) possibly belongs to 
the genus Septaliphoria, but it is not considered to be conspecific with 5. arduen- 
nensis. 

Septaliphoria paucicosta sp. n. 

(PI. 9, figs. 5-6, text-fig. 33) 

1878 Rhynchonella lacunosa (Schlotheim); Davidson: 196-97, pi. 16, figs. 13-14. 
1917 Rhynchonella helvetica (Schlotheim); Rollier: 171. 

Name. Latin paucus-a-um, few ; costa-ae, rib : the species has relatively few 
ribs. 



FROM NORTHWESTERN EUROPE 91 

Diagnosis. Medium to large Septaliphoria ; 9-15 very coarse subangular ribs ; 
no smooth area posteriorly ; uniplicate or asymmetrical ; conjunct deltidial plates ; 
strong suberect beak ; crura radulifer. 

Stratigraphical range. Oxfordian — mariae and cordatum zone ; the Lower 
Calcareous Grit of Yorkshire. 

Type specimen. Holotype, BB. 45394, selected from a box of specimens in the 
collection of the B.M.(N.H.) with the locality given as " Hutton Bushell, Beedale, 
Yorkshire ". 

Dimensions of holotype. Length 2-61 cm, thickness 178 cm, width 2-90 cm. 

Material. About 100 specimens, all from the collection of the B.M.(N.H.) 
B. 26924, B. 26903. 

Description. External characters. The width is usually greater than the 
length, with the greatest width well towards the anterior giving the species a sub- 
pentagonal to subtriangular outline. 

The beak is strongly developed in the smaller, more triangular specimens but is 
much less pronounced in the larger ones. Clearly marked beak ridges bound the 
strongly incurved interarea. An oval, submesothyridid pedicle opening is flanked 
by strongly conjunct deltidial plates. 

The shell ornament consists solely of very coarse, steep, subangular ribs ; the size 
and coarseness of the ribbing readily distinguishes the species from 5. arduennensis. 

Dimensions of Figured Specimens. 



length 


thickness 


width 


2-58 cm 


1-85 cm 


2-90 cm 


2-40 cm 


1-87 cm 


2-54 cm 



Internal characters. Pedicle valve. The dental lamellae are subparallel and 
relatively short, breaking away from the valve wall well before the insertion of the 
teeth, as seen in transverse section. The teeth are strong and there are well 
developed lateral denticulae. 

Brachial valve. A septalium is present and consists of strong septalial plates and 
and a relatively weak median septum. The sockets show only weak crenulations ; 
inner and outer socket ridges are well developed. 

The crura, which have clearly differentiated bases, are radulifer and curve towards 
the pedicle valve. At their distal ends they develop the curious processes which are 
also found in S. arduennensis. 

Distribution. As far as is known, the species has only been recorded from the 
type locality and from Filey Brigg (E. F. Owen, personal communication 1966). 

Remarks. The species only appears to have been mentioned twice in the litera- 
ture. In 1878 Davidson tentatively referred the specimen he figured from Wykham 
to Rhynchonella lacunosa Schlotheim and in 1917, Rollier suggested that Davidson's 
specimen might be R. helvetica (Schloth.). As the species does not belong to the 
" lacunosa group " and R. helvetica is here regarded as a nomen dubium, a new 
species name has been proposed for these forms from the Lower Calcareous Grit. 



9* 



JURASSIC RHYNCHONELLIDS 




Fig. 33. Transverse serial sections of Septaliphoria paucicosla sp. nov. Topotype. Oxfordian. 
Lower Calcareous Grit. Hutton Bushell, Yorks. B.M. B 26924. (X3). 



FROM NORTHWESTERN EUROPE 93 

Septaliphoria (?) hudlestoni (Rollier) 
(PI. 9, figs. 1-4) 

1878 Rhynchonella pinguis Roemer; Davidson: 193-94, pi. 16, fig. 7. 

1878 Rhynchonella pinguis var. pectunculoides Etallon; Davidson: 194-95, pi. 16, figs. 8-12. 

1 91 7 Rhynchonella Hudlestoni Rollier: 172. 

1933 Rhynchonella corallina Leymerie; Arkell: 435. 

1947 Rhynchonella pinguis Roemer; Arkell: 87. 

Emended diagnosis. Medium to large Septaliphoria-like form ; 15-20 coarse, 
subangular ribs ; uniplication in anterior commissure symmetrically or asymmetric- 
ally developed. 

Stratigraphical range. Kimmeridgian — cymodoce zone. 

Type specimen. Lectotype, here selected, fig. 10, plate 26 in Davidson (1878) ; 
collected from Abbotsbury Ironstone, Dorset. B.M. B. 27333. 

Material. 34 specimens in the collection of the B.M.(N.H.) from Abbotsbury, 
Dorset, B. 26853-54, B. 26856, B. 26858, B. 26859. 8 specimens in author's collec- 
tion, (from the same locality). 

Description. External characters. In general the specimens are very poorly 
preserved. The species appears to be very variable ; many of the specimens show 
asymmetrical development of the uniplication in the anterior commissure, although 
never the " bilobed " type of asymmetry found in Torquirhynchia. The globose 
specimens tend to show asymmetry more than the flatter ones which are often weakly 
trilobate in appearance ; the latter forms also have a higher and more pronounced 
beak. 

The beak ridges are very weak and the interarea is narrow and incurved. It would 
appear that the pedicle opening is small and circular but details of both this and the 
deltidial plates have not been satisfactorily observed. Radial ornament consists of 
about 15-20 coarse subangular ribs, while the only concentric ornament observed has 
been a few very weak growth lamellae on exceptionally well preserved specimens. 

Dimensions of Figured Specimens. 



length 


thickness 


width 


3-00 cm 


2-60 cm 


3-24 cm 


273 cm 


1-98 cm 


278 cm 


2-54 cm 


1-70 cm 


3-10 cm 



Internal characters. An attempt was made to section an apparently well pre- 
served specimen. Unfortunately, the external appearance was deceptive and 
virtually nothing could be seen of the internal structures except that a median 
septum and a possible septalium were present in the brachial valve. Mr. E. F. 
Owen (personal communication 1966) stated that he has attempted to section 
several specimens and in all cases it was impossible to determine the internal 
structure. 

Distribution. The only known occurrence of the species is in the Abbotsbury 
Ironstone of Dorset. 



04 JURASSIC RHYNCHONELLIDS 

Occurrence. According to Arkell (1933), the brachiopods are restricted to the 
" ore bed " which he described as attaining a maximum thickness of 20 feet and as 
being a " crumbling reddish-brown oolitic rock, full of shining pellets of ore ". The 
rhynchonellids are among the least abundant elements of a fauna which also includes 
the terebratulid Ornithella lampas (Sowerby), and various gasteropods, lamellibranchs 
and ammonites. 

Remarks. It has been decided to retain the specific name given by Rollier (1917) 
in preference to Rhynchonella pinguis Roemer and R. pinguis var. pectunculoides 
Etallon suggested by Davidson in view of the uncertainty attached to those names. 
If and when the latter species is revised, it may prove necessary to place the name 
hudlestoni in synonymy. This is considered unlikely, however, as the figures of 
pectunculoides given by Etallon (1861) appear to have finer ribs and a much stronger, 
more upright beak. Rhynchonella corallina Leymerie, the specific name suggested 
by Arkell, seems to be generally accepted as a synonym of R. pinguis Roemer, a 
species frequently quoted but greatly in need of elucidation and revision. 



Septaliphoria (?) septentrionalis sp. n. 
(PI. 8, figs. 6-7) 

Name. Latin, septentrionalis — northerly ; the species is only known to occur in 
the north of Scotland. 

Diagnosis. Flattened, medium sized Septaliphoria-\\ke form ; 10-15 strong, 
simple ribs ; width greater than length. 

Stratigraphical range. ? Upper Oxfordian — ? Lower Kimmeridgian. 

Type specimen. Holotype, BB. 44175, collected from Alt-na-cuile, near Brora, 
Sutherland. 

Dimensions of holotype. Length 2-37 cm, thickness 1-45 cm, width 2-86 cm. 

Material. Further 20 specimens collected by the author from Alt-na-cuile ; 
6 specimens from collection of Royal School of Mines, presumably from the same 
locality. All the specimens are preserved as sandstone casts. 

Description. External characters. The shell is either equally biconvex or with 
a slightly more inflated brachial valve. The beak is quite strong and suberect ; 
details of the beak ridges and interarea are not preserved. 

The only ornament preserved is the strong, simple ribs. Some specimens possess 
a posterior smooth area but it is not certain whether or not this is a result of the poor 
preservation. A slight uniplication was probably developed in some specimens, but 
in the best preserved material the anterior commissure tends to be asymmetrical. 

Dimensions of Figured Specimens. 

length 
2-40 cm 
2-35 cm 

Internal characters. The preservation precludes serial sectioning and only very 
limited information can be obtained from the casts. 



thickness 


width 


173 cm 


275 cm 


1-42 cm 


2-85 cm 



FROM NORTHWESTERN EUROPE 95 

The dental lamellae are strong and divergent. The median septum also appears 
to be well developed and a possible septalium is indicated by one specimen. 

Distribution. The species is only known from the type locality. 

Occurrence. The species is found in the Alt na Cuile Sandstone which Arkell 
(1933) described as consisting of " thick beds of white sandstone separated by black 
carbonaceous layers ". The associated fauna consists of lamellibranchs and 
ammonites ; the rhynchonellids are the most abundant element of the fauna. 

Remarks. According to Arkell, the Alt na Cuile Sandstone has not yet yielded 
ammonites capable of defining precisely the zones present, although it is undoubtedly 
either topmost Oxfordian or basal Kimmeridgian. As the main fossil band is 
towards the top of the unit, it seems probable that the rhynchonellids are basal 
Kimmeridgian in age. 

Arkell suggested that, as the rhynchonellids were the best preserved element of 
the fauna, it might be possible to use them to give a more accurate date to the strata. 
Unfortunately, there do not seem to be comparable forms of this age on the con- 
tinent, probably as that area was largely occupied by coral and sponge reef facies, 
with their specialized faunas, while the Alt na Cuile Sandstone almost certainly 
represents a limited near-shore facies with a fauna presumably adapted to those 
conditions. 

Although the poor preservation makes generic identification rather hazardous, the 
species is tentatively referred to Septaliphoria Leidhold. This determination is 
based on the relatively coarse, simple ribs and the fact that the anterior commissure 
may show either a symmetrical uniplication or be slightly asymmetrical ; the 
internal characters visible are consistent with this designation ; however, without 
more detailed information of the internal structures, which is unlikely to be forth- 
coming, it is not considered that a definite generic diagnosis is possible. Davidson 
(1878) suggested that the casts from Alt-na-cuile, " may also probably be referable 
to Schlotheim's species (Rhynchonella lacunosa) " . It can be safely stated that S. 
(?) septentrionallis does not belong to the genus Lacunosella. 

Genus TORQUIRHYNCHIA no v. 

1886 Inconstans-Sippe (pars) Rothpletz, p. 91. 
1918 Rhactorhynchia (pars) Buckman, pp. 50-52. 
1932 Septaliphoria (pars) Wisniewska, pp. 18-20. 

Name. Latin se torquere — to twist : all members of the genus are markedly 
asymmetrical. 

Type species. Terebratula inconstans Sowerby. 

Diagnosis. Large, asymmetrical rhynchonellids with coarse, simple ribbing ; 
septalial plates reduced or absent ; median septum present but usually very weakly 
developed ; crura radulifer. 

Stratigraphical range. In Europe the species appears to be restricted to the 
Kimmeridgian and lower Volgian ; however, Makridin (1964) has figured specimens 
from the Russian Platform which appear to belong to Torquirhynchia and which 
include material from the upper Oxfordian and the middle and upper Volgian. 



96 JURASSIC RHYNCHONELLIDS 

Description. External characters. The most characteristic feature is the strong 
asymmetry exhibited by all the species, as viewed from the anterior. The " twist " 
occurs indiscriminately either to the right or to the left. A fold and uniplication are 
never developed. Each valve has 20-30 angular or subangular ribs which are 
almost invariably simple, arising from the umbones ; however, bifurcation has 
occasionally been observed postero-laterally in T. guebhardi and T. speciosa. 

The strong beak is usually suberect although in specimens with a very inflated 
brachial valve, notably in T. inconstans, it may be adpressed to that valve. Except 
in the latter case, a functional pedicle opening is present and is flanked by disjunct to 
strongly conjunct deltidial plates. 

Internal characters. Pedicle valve. The lateral cavities and dental lamellae are 
rather variably developed, ranging between massive lamellae and correspondingly 
small lateral cavities and larger cavities bounded by slender lamellae. A slight 
pedicle collar may be developed. Strong, crenulated teeth and well developed 
lateral denticulae characterize the hinge apparatus. 

Brachial valve. A low, relatively long median septum is present ; septalial plates 
occur in all the species investigated, although they are of a distinctively reduced 
form. Inner and outer socket ridges are both well developed ; the socket floors 
are strongly crenulated. 

The crural bases differ quite markedly in their development in the different 
species described. Although the crura are radulifer in general appearance, they 
show considerable variation towards their distal ends. It is not known whether 
the development of these various distal processes can be regarded as consistent at 
species level. 

Species. The following nominal species are attributed to the genus : 

T. " astieriana " (Orbigny) (in Haas, 1891, p. 62, pi. viii, figs. 3-6, pi. ix, figs. 

i-9) 
T. " astieriana " (Orbigny) (in Makridin, 1964, pp. 107-9, P^ ™> n g- 8) 
T. " astieriana " (Orbigny) (in Wisniewska, 1932, p. 20, pi. i, figs. 21-26) 
T. astieriformis (Wisniewska) (1932, pp. 27-28, pi. i, figs. 19 ?2o) 
T. inconstans (Sowerby) (1821, vol. Ill, p. 137, pi. 277, fig. 4) 
T. guebhardi (Jacob & Fallot) (1913, p. 44, pi. v, fig. 9) 
T. pseudo-inconstans (Kitchin) (1900, pp. 51-2, pi. x, figs. 6-10) 
T. lehmanni (Makridin) (1964, pp. 109-10, pi. ii, fig. 9) 
T. pectunculoides (Etallon) (1861, p. 289, pi. xlii, fig. 3) 
T. semiconstans (Etallon) (1861, p. 290, pi. xlii, fig. 4) 
T. speciosa (Miinster) (1839, P- 1Z 3> P*- x ^i> n £- 6). 
Distribution. Although individual species appear to be relatively restricted 
geographically, the genus, taken as a whole, occurred over much of Europe west of 
the Alps. If the species of Makridin mentioned above have been correctly assigned, 
then the genus was also present on the Russian Platform. 

Occurrence. The genus appears, from the limited information available, to have 
been restricted to peri-reefal areas, and it is suggested that the asymmetrical develop- 
ment may represent an adaptation for living in a high energy environment. 



FROM NORTHWESTERN EUROPE 97 

Remarks. The genus Torquirhynchia has been proposed in order to group to- 
gether the various distinctively asymmetrical rhynchonellid species of the Upper 
Jurassic. Many of the included species have previously been attributed to Rhyncho- 
nella astieriana Orbigny. 

The Upper Jurassic, asymmetrical rhynchonellids have also been frequently 
referred to the genus Septaliphoria Leidhold, largely as a result of the paucity of 
generic names at this stratigraphical level and the lack of information regarding the 
diagnostic characters of the type species of that genus, namely S. arduennensis 
(Oppel). Internally, Torquirhynchia is readily distinguished from Septaliphoria by 
its weakly developed septalial plates and the position and development of the crural 
bases relative to them. Externally, the adult specimens of Torquirhynchia attain a 
much greater size and are always strongly asymmetrical whereas specimens of 
Septaliphoria are only weakly asymmetrical or are symmetrical. 



Torquirhynchia inconstans (J. Sowerby) 
(PI. 10, figs. 1-3, pi. 12, fig. 5, text-fig. 34) 

1821 Terebratula inconstans J. Sowerby: 137, pi. 278, fig. 4. 

1834 Terebratula inconstans J. Sowerby; Buch: 45-46. 

1838 Terebratula inconstans J. Sowerby; Buch: 146, pi. 14, fig. 16. 

1852 Rhynchonella inconstans (J. Sowerby); Davidson: 87-88, pi. 18, figs. 1-3. 

1878 Rhynchonella inconstans (J. Sowerby); Davidson: 191-93, pi. 16, figs. 1-6. 

1917 Rhynchonella inconstans (J. de C. Sowerby) (sic); Rollier: 177-78. 

1918 Rhactorhynchia inconstans (J. Sowerby); Buckman: 51. 

Emended diagnosis. Medium sized, globose Torquirhynchia ; strongly devel- 
oped beak ridges and interarea ; 25-30 simple angular ribs ; crura radulifer. 

Stratigraphical range. Kimmeridgian — baylei to cymodoce zones. 

Type specimen. Lectotype, here selected, fig. 4, pi. 277 of Sowerby (1821) ; 
the specimen is from the Kimmeridge Clay of Ringstead Bay, Dorset, in B.M.(N.H.) 
Sowerby Coll. No. B. 61475. 

Material. More than 100 specimens from the collection of the B.M.(N.H.) 
mostly from the Dorset Coast but also including a limited number of specimens from 
inland localities. 10 specimens in the author's collection obtained from Ringstead 
Bay. 

Description. External characters. The shell outline is subpentagonal and the 
valves are equally biconvex. The radial ornament consists of about 25-30 simple, 
angular or subangular ribs which show neither intercalation nor bifurcation, while 
the concentric ornament consists of numerous fine growth lines and occasional weak 
growth lamellae. In the flatter specimens the strongly developed beak is erect but 
the extent of the incurvature is obviously dependent on the globosity of the shell and 
in very globose specimens the beak is adpressed to the brachial valve. 

The relatively small hypothyridid or submesothyridid pedicle opening is flanked by 
small disjunct or conjunct deltidial plates. Beak ridges are clearly marked and 
bound the well developed interarea ; the latter may be flattened but is usually 
strongly incurved. 



98 JURASSIC RHYNCHONELLIDS 

The asymmetrical development usually commences when the animal is about half 
grown and becomes increasingly pronounced with age. 

Dimensions of Figured Specimens. 

length thickness width 

3-20 cm 3-10 cm 3-25 cm 

3-02 cm 2-64 cm 3-12 cm 

3-20 cm 2-25 cm 3-31 cm 

Internal characters. Pedicle valve. The shell wall is very thick and the small 
lateral cavities are not seen in transverse section until about the level of the top of 
the brachial valve. The dental lamellae are massive and only attached to the 
ventral shell wall at their extreme posterior ends. The teeth are strong and heavily 
crenulated ; the general impression given by the teeth, dental lamellae and the way 
the pedicle valve abuts against the brachial valve in the earlier sections, is that the 
species possessed a very strongly articulated shell. 

Brachial valve. The sockets are crenulated and there are both inner and outer 
socket ridges. A median septum is present but it does not come into contact with the 
septalial plates as it is only present posteriorly as a low ridge. The septum com- 
pletely disappears, as seen in transverse section, before the end of the crura is 
reached, but shows quite a strong development for part of its length. The distinctive 
short, rounded septalial plates are seen on the dorsal side of the ventrally curved 
hinge plates. Crural bases are hardly differentiated. At first, as seen in transverse 
section, the radulifer crura appear as structures flattened in the plane of articulation, 
but in later sections they become more rod-like and finally become flattened at right 
angles to the plane of articulation and curve towards the pedicle valve. 

Distribution. The best known localities are those of Ringstead Bay and 
Weymouth in Dorset ; inland, it has been recorded from Shotover Hill near Oxford, 
Wooton Basset in Wiltshire, Brill in Buckinghamshire and from Swindon. On the 
continent as observed by Haas (1890), the species appears to be restricted to north- 
western France. Buch (1838) recorded it from Ellrichserbring in Brunswick. His 
figure of the species is a copy of Sowerby's (1821) figure 4 and, as he did not illustrate 
any material from his locality, the record must be regarded as doubtful. 

Occurrence. At Ringstead Bay it occurs in a thin bed of pale grey clay, lying 
between the Ringstead Coral Bed and the Exogyra Bed, as illustrated in Arkell 
(1933, pi. 21). The associated fauna includes terebratulids, the gasteropod Bathroto- 
maria reticulata (Sowerby) and ammonites. Ager (1965a) has drawn attention to 
the probability of T. inconstans living in a peri-reefal environment, as evidenced by 
the close proximity of the Coral Bed. At Shotover Hill it is recorded by Phillips 
(1855) as occurring in a band of septarian nodules in a shaly clay of cymodoce age. 

Remarks. In his discussion of the species, Davidson (1852) included a specimen 
(pi. 18, fig. 4) from the Inferior Oolite of Leckhampton Hill. The author has not 
seen the actual specimen, but assumes this to be a misidentification in view of the 
otherwise restricted record of the species. On the same plate fig. 3, Davidson 
figured a symmetrical specimen from Shotover Hill, again the author has not seen the 



FROM NORTHWESTERN EUROPE 



99 




100 



Fig. 34. Transverse serial sections of Torquirhynchia inconstans (Sowerby). Topotype. 
Kimmeridgian. Kimmeridge Clay. Ringstead Bay, Dorset. (X2). 



ioo JURASSIC RHYNCHONELLIDS 

specimen but it seems unlikely that it is correctly determined as, apart from its 
symmetry, the ribbing and general shape also differ from inconstans. The B.M. 
(N.H.) collection contains only one symmetrical specimen attributed to inconstans, 
a large trilobate form again from Shotover Hill. The horizon of this specimen is not 
known but, despite crushing, it is interesting in that it appears closer to Somalir- 
hynchia sutherlandi than to T. inconstans. 

While collecting material at Ringstead Bay, it was obvious that a fairly high 
proportion of the specimens obtained were unusual in that the valves " gaped". 
The author has since examined the specimens available at the B.M.(N.H.), about 
ioo, and this has confirmed that about a third show this feature. Many specimens 
have a well developed epifauna and the tubes of ?Serfinla are frequently seen on 
both valves. 



Torquirhynchia guebhardi (Jacob & Fallot) 
(PI. io, figs. 4-6, text-fig. 35) 

1913 Rhynchonella Astieviana Orbigny var. Guebhardi Jacob & Fallot: 45-46, pi. 5, fig. 9. 
191 7 Rhynchonella Guebhardi Rollier: 77. 

Emended diagnosis. Medium-sized, globose Torquirhynchia ; markedly asym- 
metrical with a frontal notch between the two lobes ; about 30 subangular ribs ; 
radulifer crura flattened in the plane of articulation. 

Stratigraphical range. Jacob & Fallot describe it as coming from the 
" Portlandien "; the author has collected it from the cymodoce zone of the Kim- 
meridgian. 

Type specimen. Lectotype, here selected, the specimen figured by Jacob & 
Fallot (1913), pi. 5, fig. 9 ; the type locality is St. Vallier (Alpes-Maritimes). 

Material. 7 specimens from a road cutting north of Lac d'Armaille (Ain) in 
British Museum (Nat. Hist.) nos. BB. 45762-68. Other localities in the southern 
Jura from which the species has been collected include Rossillon, Glandieu, Brognin 
Hill and Chavoley. 

Description. External characters. The valves are almost equally biconvex, 
with a tendency for the brachial to be slightly the more inflated. However, the 
valves are distinctively very globose and this, coupled with the very strongly 
developed asymmetry, differentiates T. guebhardi from other species in the genus. 
The beak is sharp and suberect and the shell outline subtriangular. 

T. guebhardi possesses about 30 relatively fine, subangular ribs which arise at the 
umbones and continue simply to the anterior margin ; bifurcation has only been 
observed in a few postero-laterally positioned ribs. The concentric ornament con- 
sists only of rather sparse, weak growth lamellae. 

Although the beak is quite large and suberect, the beak ridges are only weakly 
developed as are the small, incurved interareas. An oval, submesothyridid pedicle 
opening is flanked by conjunct deltidial plates. In transverse section the deltidial 
plates are clearly seen to project externally around the foramen. 



from northwestern europe 101 

Dimensions of Figured Specimens. 

length thickness width 

2-97 cm 2-20 cm 3-35 cm 

2-55 cm 1-95 cm 280 cm 

Internal characters. Most of the fine detail of the structures has been lost through 
recrystallisation. 

Pedicle valve. The narrow, elongate lateral cavities, which penetrate almost to 
to the apex of the beak, are limited by slender dental lamellae. A pedicle collar is 
present. The teeth are strong and crenulated ; lateral denticulae are present but 
not strongly developed. 

Brachial valve. There is no septalium and only a weak median septum ; small 
septalial plates are present. The sockets are crenulated ; inner and outer socket 
ridges well developed. Crural bases are not well differentiated. The radulifer 
crura are flattened in the plane of articulation and do not curve towards the pedicle 
valve. 

Distribution. The species only appears to have been recorded from the following 
two areas. (1) The material of Jacob & Fallot (1913) came from St. Vallier in the 
French Alpes Maritimes and in the " Portlandien coralligene des environs de Gerin, 
Ain ". The latter name is probably a mis-spelling of Cerin. (2) All the material 
studied by the author was collected by various members of Imperial College from the 
southern French Jura in the area around Belley, which is also in the Department of 
Ain. 

Occurrence. The material described by the author was collected from the 
Bedded Virieu Limestone (Ager and Evamy, 1963). The following description of 
mode of occurrence and lithology is taken from Evamy (1963), " The subdivision 
(of the Bedded Virieu containing T. guebhardi) is about 20 m thick — it consists 
mainly of alternating limestones and shales, which yield an abundant brachiopod, 
lamellibranch fauna, as well as ammonites of the tenuilobatum and pseudomutabilis 
zones — . The limestones are thinly bedded (20-40 cm) brown calcilutites, occasion- 
ally showing a mottled iron-staining. These are separated by thin (approximately 
10 cm) shaly bands." He also noted that, " A few thicker beds of calcilutite 
(i-i| m) are seen not to contain the abundant mollusc and brachiopod fauna ." 
The beds containing T. guebhardi immediately underlie the reef horizon of the 
Massive Virieu Limestone. 

Remarks. Although the author has not seen the original material of Rhyncho- 
nella Astieriana var. Guebhardi, it is thought that a conspecific identification for the 
material from the Bedded Virieu Limestone can confidently be given as it matches 
exactly the specimen figured by Jacob & Fallot (1913). Also, the Bedded Virieu 
could well be the lateral stratigraphical equivalent of the " Portlandien " of Jacob 
and Fallot. Referring to the Bedded Virieu, Ager and Evamy (1963) note that it 
is, " strongly reminiscent of the ' Tithonien ' facies seen in many parts of Europe 
at this level ". In the latter publication, T. guebhardi is referred to as Septaliphoria 
astieriana (d'Orbigny). 



JURASSIC RHYNCHONELLIDS 




3 8 



3 5 



3 



2-7 



2 6 





4 7 




5 3 




5 8 



Fig. 35. Transverse serial sections of Torquirhynchia guebhardi (Jacob & Fallot). Kim- 
meridgian — cymodoce zone. N. of Lac d'Armaile (Ain), France. ( X3). 



FROM NORTHWESTERN EUROPE 103 

Torquirhynchia cf. T. astierformis (Wisniewska) 
(PI. 12, figs. 1-3, text-fig. 36) 

1932 Septaliphoria pinguis (Roemer) var. astierformis Wisniewska: 27-28, pi. 1, fig. 19, ?2o. 

Emended diagnosis. Relatively flattened, medium sized Torquirhynchia ; 
about 20 subangular ribs ; beak small ; asymmetrical development not sufficient 
to produce a frontal notch ; crura radulifer. 

Stratigraphical range. The material described by the author came from the 
mutabilis zone of the Kimmeridgian, while the specimens figured by Wisniewska 
came from the " Rauracien sup." and " Kimmeridgien ? 

Material. 6 specimens collected by the author from Sermerieu (Isere), including 
B.M. BB.45174, BB.45175, BB.45176. 

Description. External characters. The width is greater than the length and 
the greatest width is towards the anterior ; this, combined with the rather steep 
apical angle, results in an overall subtriangular outline. T. cf. T. astieriformis is 
equally biconvex and relatively flattened as compared with other species of the 
genus. The coarse ribs arise at the umbones and continue simply to the anterior 
margin ; concentric ornament has not been observed. 

The beak is suberect and flanked by slight beak ridges which delimit the weakly 
developed and slightly incurved interarea. Disjunct deltidial plates define the 
relatively small pedicle opening. Asymmetry is characteristic but is not so marked 
as in T. guebhardi as it does not result in the development of a frontal notch. 

Dimensions of Figured Specimens. 

length thickness width 

3-18 cm 255 cm 3-49 cm 

2-63 cm 2-94 cm 3-00 cm 

2-32 cm 247 cm 273 cm 

Internal characters. Pedicle valve. The lateral cavities are bounded by strong, 
subparallel dental lamellae which break away from the ventral wall of the valve, as 
seen in transverse section, before the insertion of the teeth. The teeth are strong 
and crenulated ; well developed lateral denticulae are present. 

Brachial valve. The sockets show tegulate crenulations and both inner and outer 
socket ridges are strongly developed. The redulifer crura curve towards the pedicle 
valve ; at their distal ends they produce the rather peculiar processes seen in 
text-fig. 36. 

Distribution. As stated above, the author's material came from Isere where, 
according to Dr. Enay (personal communication 1964), it is not an uncommon 
species. The only figured material likely to be conspecific is that of Wisniewska 
(1932) from Poland. 

Occurrence. The species was collected from irregularly interbedded, fine 
grained limestones and marls. The exposure consisted of a small roadside cutting 
and the varied fauna present was largely collected from around a small sponge reef 
about 1 m high and 2 m across. Apart from the rhynchonellids, the fauna included 



io4 



JURASSIC RHYNCHONELLIDS 




70 



Fig. 36. Transverse serial sections of Torquirhynchia cf. — aslieriformis (Wisniewska). 
Kimmeridgian — mutabilis zone. Serm6rieu (Isere), France. (X3). 



FROM NORTHWESTERN EUROPE 105 

crinoid ossicles, zeilleriids and terebratulids as well as sponges. The rhynchonellids 
were interesting in that the two species present belonged to two different genera 
namely, Tor quirky nchia and Lacunosella ; both species were associated with the 
sponges and Dr. Enay stated that this is a characteristic association of the mutabilis 
zone in that area. T. cf. astieriformis was the least abundant element of the 
fauna. 

Remarks. Although externally one, at least, of the specimens figured by 
Wisniewska very closely resembles the material from Isere, the lack of records from 
the intervening area and the lack of knowledge as to the internal structures of the 
Polish material precludes a definite conspecific identification of the French and 
Polish specimens. 

T or quirhy nchia speciosa (Miinster) 
(PI. ii, figs. 1-3, text-fig. Z7) 

1832 Terebratula diffovmis Lamarck; Zieten: 56, pi. 62, figs. 2. 

1839 Terebratula inconstans speciosa Miinster: 113, pi. 13, fig. 6. 

1847 Rhynchonella Astieriana Orbigny: 14-15, pi. 492, figs. 1-4. 

1850 Rhynchonella Astieriana Orbigny; Orbigny: 24. 

1852 Terebratula inconstans Quenstedt: 455, pi. 36, fig. 44. 

1858 Rhynchonella Astieriana Orbigny; Suess: 52, pi. 6, fig. 2. 

1863 Rhynchonella inconstans Orbigny; Ooster: 47, pi. 15, figs. 1-13. 

1871 Terebratula inconstans speciosa Quenstedt: 138-39, pi. 15, figs. 45, 51, ?5o. 

1885 Terebratula inconstans Quenstedt: 694, pi. 53, fig. 64. 

1913 Rhynchonella Astieriana Orbigny; Jacob & Fallot: 43, pi. 5, 

1917 Rhynchonella Astieriana Orbigny; Rollier: 176. 

191 7 Rhynchonella speciosa Rollier: 178. 

Emended diagnosis. Largest known species of Tor quirhy nchia, being up to 
10 cm wide ; 20-30 coarse, angular ribs ; large prominent suberect beak ; asym- 
metrical ; width always greater than length ; crural bases sharply pointed ; crura 
radulifer. 

Stratigraphical range. Kimmeridgian — lower Volgian : limestones in which 
topotypes occur range from subeumela to Gravesia zones (Dr. Barthel, personal 
communication 1965). 

Type specimen. Lectotype, here selected, is the specimen originally figured by 
Miinster, pi. 13, fig. 6 ; a plaster cast of this is in the collection of the B.M.(N.H.), 

B. 5513- 

Dimensions of lectotype. Taken from plaster cast mentioned above : length 
5-22 cm, thickness 3-35 cm, width 7-92 cm. 

Material. 28 specimens in the collection of the author including B.M. BB.45171, 
BB.45173 from Saal, Germany. 2 specimens from collection of the University of 
Tubingen. 

Description. External characters. T. speciosa is one of the largest Mesozoic 
rhynchonellids : Quenstedt (1871) recorded a specimen over 10 -o cm wide. The 
species shows considerable variation in outline but, as the greatest width is well 
toward the anterior, it is usually subtriangular or suboval. There are about 20-30 



io6 



JURASSIC RHVNCHONELLIDS 



very strong, sharply angular ribs which arise from the umbones and continue simply 
to the anterior margin. Concentric ornament, not always preserved, consists of 
numerous fine growth lines, together with occasional coarser growth lamellae. 

The species has a characteristically strong, high, suberect beak which is flanked by 
a large, well-marked interarea, although the beak ridges limiting these are only 
weakly developed. T. speciosa possesses a large hypothyridid pedicle opening. 

Dimensions of Figured Specimens. 

length thickness width 

5-50 cm 2-30 cm 687 cm 

473 cm 2-42 cm 5-00 cm 

260 cm 1-70 cm 3-05 cm 




4 6 



3 4 




6 5 

Fig. 37. Transverse serial sections of Torqidrhynchia speciosa (Munster). Kimmeridgian/ 
Lower Volgian. " Diceras Kalk ". Saal, Germany. (X3). 

Internal characters. Pedicle valve. There is a weakly developed pedicle collar. 
The deltidial plates project slightly, " reinforcing " the pedicle opening ; below the 
level of the pedicle opening they become conjunct and then " crumple up " against 
one another. Relatively slender, subparallel dental lamellae are present. The 
teeth are strong and crenulated ; there are well developed lateral denticulae. 



FROM NORTHWESTERN EUROPE 107 

Brachial valve. The median septum is weak and there are no septalial plates. 
The crural bases are clearly differentiated. The radulifer crura are relatively small 
and curve only slightly to the ventral of the plane of articulation. The sockets show 
tegulate crenulations ; inner and outer socket ridges are both strongly developed. 

Distribution. Judging from the figures and descriptions in the literature, the 
species has a rather discontinuous distribution. The material figured by Zieten 
(1832), Miinster (1839) an d Quenstedt (1852, 1871, 1885) came from the Franconian 
and Swabian Jura, while the type localities for Rhynchonella Astieriana of Orbigny 
(1847) are m the French " Alpes Maritimes " at Escragnolles and La Malle. Other 
figured specimens which would appear to be conspecific are from " Les Alpes 
Bernoises et Vaudoises " (Ooster, 1863), and from Stramberk, Czechoslovakia, and 
adjacent areas of Poland (Suess, 1858). 

Occurrence. At Saal, T. speciosa is associated with a very varied fauna which 
includes many large species, or at least large specimens, of terebratulids, gasteropods 
and lamellibranchs. One of the most distinctive elements here is the lamellibranch 
Diceras, which gives its name to the limestone. Dr. Barthel (personal communica- 
tion 1965) suggests that the Diceras-Kalk, as seen at Saal and Kelheim, represents a 
fore-reef limestone. Unfortunately, the details at Saal are largely obscured by the 
recrystallization, tectonics and fissure infillings of younger material. 

Remarks. Although the name Rhynchonella Astieriana of Orbigny (1847) has 
been very widely used and is embedded in the literature, it is an indisputable object- 
ive junior synonym of Terebratula inconstans speciosa of Miinster (1839). ^ n hi s 
original description of R. Astieriana Orbigny gave Miinster 's species in his synonymy 
and as, under the present rules of nomenclature, it is not permitted to arbitrarily 
replace a specific or subspecific name with another, Miinster's name speciosa must be 
given priority. Since Orbigny, it appears that only Quenstedt (1871) has described 
Astieriana as a synonym of speciosa while other authors such as Suess (1858) and 
Haas (1870) have either retained the name Astieriana while admitting speciosa as a 
synonym of it, or ignored speciosa altogether, e.g., Jacob and Fallot (1913). In 
general, the name Astieriana has tended to be used for all rhynchonellids in the 
Upper Jurassic showing asymmetry. 

Terebratula difformis of Zieten is a junior homonym of T. diffomiis of Lamarck. 

"Rhynchonella" ordinaria sp. n. 

(PL 9, figs. 7-1 1, text-figs. 38-39) 

Name. Latin ordinarius-a-um, ordinary ; the species is a very " ordinary look- 
ing " rhynchonellid. 

Diagnosis. Medium sized rhynchonellid ; strong suberect beak ; 15-20 simple, 
subangular ribs ; there is a weak median septum ; septalial plates present but only 
poorly developed ; crura radulifer. 

Stratigraphical range. Of the two occurrences of the species known one is in 
the pseudocordata zone of the Oxfordian and the other is in the " Sequanien ". 

Type specimen. Holotype, BB. 45167, was obtained from the pseudocordata 
zone at Mont Dolet near Sermerieu (Isere), France. 



108 JURASSIC RHYNCHONELLIDS 

Dimensions of holotype. Length 2-50 cm, thickness 178 cm, width 2-20 cm. 

Material. 20 specimens collected by the author from the type locality, British 
Museum (Nat. Hist.) nos. BB. 45710-19, BB. 45770-79 ; 30 specimens from near 
Bourges (Cher) collected by M. Delance, (University of Dijon). 

Description. External characters. The species is an ordinary looking rhyn- 
chonellid with the valves either equally biconvex or with the brachial valve slightly 
the more inflated. Length and width are about equal and either may be the greater ; 
maximum width may be well toward the anterior or roughly median, thus giving a 
considerable range of overall outline. The young specimens are more flattened and 
subtriangular. 

Most specimens have a high, strong beak which is flanked by strong beak ridges. 
The interarea is markedly incurved and is developed most strongly at the base of the 
beak where it impinges on the brachial valve giving a distinctive bend to the hinge 
line. A large, circular, hypothyridid pedicle opening is present and is limited by 
conjunct deltidial plates. 

The radial ornament consists of rather coarse, steep, simple ribbing, while the 
concentric ornament comprises numerous very fine growth lines. The anterior 
commissure either shows a symmetrical low uniplication or varying degrees of 
asymmetry ; the latter feature is never so strongly developed as in Torquirhynchia, 
but shows the same sort of variation as described for Septaliphoria. 

Dimensions of Figured Specimens. 

length thickness width 

126 cm 0-51 cm 1-25 cm 

1-82 cm 077 cm 166 cm 

2-40 cm 1-85 cm 2-17 cm 

2-28 cm 1-42 cm 2-18 cm 

250 cm 178 cm 2-20 cm 

Internal characters. Pedicle valve. The beak is massive and the sub-parallel 
dental lamellae are quite short resulting in the development of only small lateral 
cavities. A small pedicle collar is present. The teeth are strong and crenulated and 
the species also possesses strong inner and outer socket ridges. 

Brachial valve. The hinge plates are arched ventrally and the inner hinge plate 
is distinctive in showing a dorsally directed ridge flanked by small septalial plates. 
Crural bases are differentiated ; the relatively small radulifer crura curve slightly 
toward the ventral valve. 

Distribution. Only known from the type locality and from near Bourges (Cher), 
France. 

Occurrence. The specimens from Isere were collected from a band of fine- 
grained limestone, about 1-25 m thick, in which they occurred as loose groups. The 
associated fauna consisted of occasional zeilleriids and lamellibranchs, including 
oysters. M. Delance, of the University of Dijon, describes his material as coming 
from a " calcaire crayeux blanc, tres friable ", which lies above a massive reef lime- 
stone containing corals. 



FROM NORTHWESTERN EUROPE 



109 




6 9 



Fig. 38. Transverse serial sections of " Rhynchonella " ordinaria sp. nov. Topotype. 
Oxfordian — psendocordata zone. Mont Dolet (Isere), France. (X3). 



JURASSIC RHYNCHONELLIDS 




7 2 



7 8 



Fig. 39. Transverse serial sections of " Rhynchonella " ordinaria sp. nov. " Sequanien 

Bourges (Cher), France. (X4). 



FROM NORTHWESTERN EUROPE in 

Remarks. Externally, the species appears similar to forms such as " R." 
hopkinsi of the Bathonian which are normally referred to Burmirhynchia Buckman. 
However, there is considerable doubt as to many of the features of that genus and 
nobody appears to have published serial sections of topotypic material. Mr. E. F. 
Owen has kindly allowed the author to see sections of a Burmese specimen in the 
reference collection of the B.M.(N.H.), and while those seem to be closely comparable 
with those made of " R." ordinaria, this in itself is somewhat odd in view of the 
original description of the genus in which Buckman stressed the prominent nature of 
the median septum and dental lamellae, features only poorly developed in " R." 
ordinaria. These latter features can be clearly seen in the burnt specimens figured 
by Buckman (1918) and are also in a specimen attributed to R. hopkinsi from 
Chatillon-sur-Seine, which was sectioned for comparison. Although it is very 
difficult to judge to what extent such features observed externally will be developed 
as seen in transverse section, it is considered that their very weak development in 
" R." ordinaria represents a marked discrepancy as compared with the species 
figured as Burmirhynchia by Buckman. 

In general shape and form of ribbing, " R." ordinaria is also comparable with 
certain species of the genus Kallirhynchia Buckman. Internally, it shows some 
resemblance to Kallirhynchia yaxleyensis, as figured by Muir-Wood (1934), especially 
in the development of its septalial plates and in the shape of the crura. However, 
Kallirhynchia, as it is at present defined, is restricted to the Bathonian and con- 
sequently it seems unwise to place ordinaria in that genus until intermediate forms 
have been described. At present the rhynchonellid faunas of the Bathonian and 
Callovian are very imperfectly known. 



"Rhynchonella" pyrenaei sp. n. 

(PI. 3, fig. 7, text-fig. 40) 

Name. Latin — Pyrenaeum-i, the Pyrenees ; the type locality is near Pau 
(Basses-Pyrenees) . 

Diagnosis. Small to medium sized rhynchonellid ; length greater than width ; 
beak relatively large and suberect ; about twelve ribs, some of which bifurcate ; 
low uniplication ; crura radulifer. 

Stratigraphical range. Lower Oxfordian. 

Type specimen. Holotype from the Pic du Gar near Pau (Basses-Pyrenees). 
BB. 45161. 

Material. Large number of broken and poorly preserved specimens in the 
author's collection. 

Dimensions of holotype. Length 1-51 cm, thickness 0-87 cm, width 1-29 cm ; 
this is the only specimen figured, B.M. BB.45161. 

Description. External characters. The beak ridges and interarea are only 
weakly developed. The material was not sufficiently well preserved to allow the 
elucidation of the details of the pedicle opening and the deltidial plates. Concentric 
ornament was only observed on one specimen and consisted of growth lamellae. 



112 



JURASSIC RHYNCHONELLIDS 




3 2 



3 9 



Fig. 40. Transverse serial sections of " Rhynchonella " pyrenaei sp. nov. Topotype. 
Oxfordian. Pic du Gar, (Basses Pyrenees), France. (X5). 



Internal characters. Pedicle valve. The delthyrial cavity is subquadrate and 
limited by long, strong dental lamellae. Inner socket ridges are well developed ; 
lateral denticulae are present. 

Brachial valve. The median septum is strong and long septalial plates are 
developed. As seen in transverse section, the crura possess a very distinctive 
appearance through having " knob-like " structures median of the sharply pointed 
crural bases. 

Distribution. Only known from the type locality. 

Occurrence. All the material came from a bed about 10 cm thick where the 
species occurred profusely together with a lesser number of terebratulids. The 
matrix consisted of a very hard, dark limestone from which it was very difficult to 
extract specimens. 



FROM NORTHWESTERN EUROPE 113 

Remarks. The species has not been referred to a genus as it is markedly dissimilar 
to any other species known to the author occurring in the Oxfordian. The dis- 
tinctively shaped crura are closely comparable to those of Prionorhynchia serrata (J. 
de C. Sowerby) figured by Ager (1956) ; however, there is little resemblance extern- 
ally and the latter is only known from the Lower Jurassic. 

ACKNOWLEDGMENTS 

The writer would like to record his gratitude to the many people who have assisted 
him during the course of this study. He is particularly grateful to Dr. D. V. Ager 
who originally suggested the main topic for research and who has subsequently given 
much valuable advice and guidance. 

The writer is also indebted to those museums, universities and individuals who 
have loaned and donated material from their collections. These include Dr. Barthel, 
Bayerische Staatsammlung, Munich ; M. Delfaud, University of Bordeaux ; 
Dr. Dreyfuss, University of Besancon ; Dr. Enay, University of Lyon ; Dr. Gidon, 
University of Chambery ; H. Panchaud and Dr. Gasche, Naturhistorisches Museum, 
Basel ; Dr. Rioult, University of Caen ; Dr. Schumann, University of Tubingen ; 
Prof. Tintant and M. Delance, University of Dijon and Dr. Zeiss, University of 
Erlangen. Special thanks are due to the representatives of the Societe Nationale 
des Petroles d'Aquitaine, particularly M. Capdecomme, and to Dr. Maubeurge for 
assistance in the field in the Pyrenees and the area around Nancy respectively. With 
regard to loans of British material, the writer particularly wishes to thank Mr. E. F. 
Owen of the British Museum (Natural History) ; Mr. J. M. Edmonds, University 
Museum, Oxford ; Mr. A. G. Brighton, Sedgewick Museum, Cambridge and Dr. R. B. 
Wilson of the Geological Survey, Edinburgh. 

Mr. J. A. Gee gave much advice and assistance in photographic matters for which 
the author is very grateful. 

The author also wishes to acknowledge the useful discussions with his colleagues 
Dr. P. Copper and Dr. D. A. B. Pearson and he is also indebted to the latter for 
donating a large collection of Polish specimens. 

Grateful thanks are recorded to the author's wife for continual support and 
encouragement as well as for valuable assistance during fieldwork. 

Finally the writer would like to thank the Natural Environment Research Council 
for the award of a Research Studentship allowing him to carry out this study at 
Imperial College, London, and also the Principal and Governors of The College 
of Technology, Oxford, for providing the facilities to complete the manuscript. 



ii 4 J I RASSIC RHYNCHONELLIDS 

REFERENCES 

A.GER, I). V. 1956. A Monograph of the British Liassic Rhynchonellidae. Palaeontogv. Soc. 

[Monogr.] J., London, (1): 1-50, pi. 1-4. 
1957- The true Rhynchonella. Palaeontology, London, (1), 1 : 1-15, pi. 1-2, 4 figs. 

- 1959. Lower Jurassic brachiopods from Turkey. /. Paleont., Chicago, 33 : 1018-28. 
i960. Nomenclatural Problems in the Mesozoic Rhynchonelloidea. Geol. Mag., London, 

97 : (2), 157-62. 

- 1962. A Monograph of the British Liassic Rhynchonellidae. Palaeontogv. Soc. [Monogr.], 
London, (3) : 85-136, pi. 8-1 1. 

- 1964. The British Mesozoic Committee. Nature, London, 203, no. 4949 : 1059. 

- 1965a. The Adaptation of Mesozoic Brachiopods to Different Environments. Palaeo- 
geography, Palaeoclimatol, Palaeoecol, 1, 143 — 172. 

- 19656. Mesozoic and Cenozoic Rhynchonellacea. In : R. C. Moore (Editor) Treatise on 
Invertebrate Paleontology, H. Kansas. 

Acer, U. V. & Evamy, B. D. 1963. The Geology of the Southern French Jura. Proc. Geol. 

Ass., London, 74, (3) : 325-56, figs. 1-5, 2 maps. 
Arcelin, F. & Roche, P. 1936. Les Brachiopodes bajociens du Monsard. Trav. Lab. G60I. 

Univ. Lyon., fasc. 30, (Mem 25) : 5-107, pi. 1-19, figs. 1-29. 
Arkell, W. J. 1933. The Jurassic System in Great Britain. Clarendon Press, Oxford, i-xii. 

+ 1-681, pi. 1-41. 

1947. The Geology of the Country around Weymouth, Swanage, Corfe and Lulworth. 

Mem. geol. Surv. U.K., London, i-xii +1-386, pi. 19, 84 figs. 

Bailey, E. B. & Weir, J. 1932. Submarine Faulting in Kimmeridgian Times. Trans, roy. 

Soc. Edinb., 57 : 429-67. 
Blake, J. F. 1880. On the Portland Rocks of England. Q. Jl geol. Soc. Lond., 36 : 189-236, 

pi. 9-10. 
Boye, N. 1842-3. Importance de l'etude des fossiles pour la reconnaissance geologique des 

terrains. Mem. C.R. Soc. Emul. Doubs. (3), 1 : 13-17, 4 pi. 
Branco, W. 1879. Der untere Dogger Deutsch-Lothringens. Abh. geol. Spezkarte Els. -Loth., 

2, (1) : i-vi +1-160, 10 pis. 
Buch, L. von. 1834. Ueber Terebrateln. Abh. dt. Akad. Wiss. Berlin, (for 1835) : 21-144, 

pi. 1-3. 

1838. Essai d'une Classification et d'une Description des Terebratules. M6m. Soc. giol. 

Fr., Paris, (1), 3 : 107-238, pi. 13-20. 

Buckman, S. S. 1883. The Brachiopoda from the Inferior Oolite of Dorset and a portion of 

Somerset. Proc. Dorset nat. Hist, antiq. Fid. Club, Dorchester, 4 : 1-52. 

1914- Genera of the Jurassic Brachiopoda. 2 pp. London. 

1918. The Brachiopoda of the Namyau Beds, Northern Shan States, Burma. Mem. geol. 

Surv. India Palaeont. indica, Calcutta. New Series, 3, (2) : 1-299, 2I pis- 
Buckman, S. S. & Walker, J. F. 1889. On the spinose Rhynchonellae (Genus Acanthothyris, 

d'Orbigny), found in England. Rep. Yorks. phil. Soc, York (for 1888) : 41-57. 
Chiplonker, G. W. 1938. Rhynchonellids from the Bagh Beds. Proc. Indian Acad. Sci. (B), 

7, (6) : 300-16, 1 pi., 7 figs. 
Croft, W. N. 1950. A parallel grinding instrument for the investigation of fossils by serial 

sections. J . Paleont., Chicago, 24, (6) : 693-8, 4 figs. 
Dagis, A. S. 1963. Upper Triassic brachiopods of the southern U.S.S.R. Akad. Nauh. 

S.S.S.R. Moskow. 1-248, pi. 1— 31. (in Russian). 
Dall, W. H. 1877. Index to the Names which have been applied to the Subdivisions of the 

Class Brachiopoda. Bull. U.S. natn. Mus., Washington, 8 : 1-88. 
Davidson, T. 1851-2. A Monograph of the British Fossil Brachiopoda. r, (3), The Oolitic 

and Liassic Brachiopoda. Palaeontogr. Soc. [Monogr.], London, 1-64 : pi. 1-13, 1851 ; 65- 

100, pi. 14-18, 1852. 



FROM NORTHWESTERN EUROPE 115 

Davidson, T. 1855. Ibid. 1, Appendix and supplementary Additions to the Appendix Pala- 

eontogr. Soc. [Monogr.], London, 1-30, pi. A. 
1873. Note on some Brachiopoda collected by Mr. Judd from the Jurassic Deposits of the 

East Coast of Scotland. Q. Jl geol. Soc. Lond., 29 : 196-7, pi. 8. (Appended to Judd). 

1874-82. A Monograph of the British Fossil Brachiopoda. 4, (1), Tertiary, Cretaceous, 

Jurassic, Permian and Carboniferous Supplement and Devonian and Silurian that occur in 
the Trias pebble bed of Budleigh Salterton in Devonshire. Palaeontogr. Soc. [Monogr.], 
London, 1-383, 42 pi. 

1876-8. Ibid. 4, (2), Supplement to the British Jurassic and Triassic Brachiopoda. 

Palaeontogr. Soc. [Monogr.], London, 73-144, pi. 10-16, 1876; 145-242, pi. 27-9, 1878. 

1877. On the Species of Brachiopoda that occur in the Inferior Oolite of Bradford Abbas 

and its Vicinity. Proc. Dorset nat. Hist, antiq. Fid Club, 1 : 73-88, pi. 1-4. 
1884. A Monograph of British Fossil Brachiopoda. 5, (3), Appendix to the Supplements 

and General Summary. Palaeontogr. Soc. [Monogr.], London, 243-399. pi. 18-2 1. 
Defrance, M. J. L. 1828. Tiribratules fossiles. In : Dictionaire des Sciences Naturelles. 

53 : 127-67, Paris. 
Deslongchamps, E. E. 1857. Descriptions des couches du systeme oolithique inferieur du 

Calvados. Bull. Soc. linn. Normandie, Caen, 2 : 312-67, pi. 4-5, 1 table. 
1859. Notes sur le terrain Callovien. 2 Note sur le Callovien des environs d'Argentan, et de 

divers points du Calvados. Bull. Soc. linn. Normandie, Caen, 4 : 1 1-16, pi. 4, 1 fig., 1 table. 

1863-87. Etudes critiques sur des Brachiopodes nouveaux ou peu connus. Bull. Soc. 

linn. Normandie, Caen, (3), 8 : 161-350, pi. 1-14, 1884 ; (3), 10 : 31-158, pi. 27-8, 1887. 

Elliott, G. F. 1950. The Genus Hamptonina (Brachiopoda) ; and the relation of post 

palaeozoic brachiopods to coral reefs. Ann. Mag. nat. Hist., London, (12), 3 : 429-47, 1 pi. 
Etallon, A. 1861. Etudes paleontologique sur les terrains Jurassique du Haut Jura. Mono- 
graph de l'etage corallien. Mem. Soc. Emul. Doubs. (3), 6, (1) : 53-216. 
Evamy, B. D. 1963. The Geology of the French Jura Mountains West of Belley (Ain). Ph.D. 

thesis, Univ. Lond. 
Fischer de Waldheim, G. 1809. Notice des Fossils du Gouvernement de Moscou, pt. i, sur 

les coquilles fossiles dites Terebratules. Invitation a la Seance Publique do la Sociite Im- 

piriale des Naturalistes de Moscou. 1-35, pi. 1-3. (Moscow). 
Fraas, O. 1858. Geognostische Horizonte im weissen Jura. Jh. Ver. vaterl. Naturk. Wiirtt. 

1 : 97-U4- 
Gerasimov, P. A. 1955. Guide to the Mesozoic fossils of the central region of the European 

part of the U.S.S.R. Part 1, Lamellibranch, Gasteropod, Scaphapod, Mollusca, and 

Brachiopoda in the deposits of the region. State Scientific Technical Publications on Geology 

and Natural Resources, Moscow. 1-379, pi- 1-50 (in Russian). 
Geyer, G. 1889. Uber die Liasischen Brachiopoden des Hierlatz bei Hallstatt. Abh. geol. 

Bundesanst. Wien. 15 : 1-88, pi. 1-9. 
Gillieron, V. 1873. Alpes de Fribourg en general et de Montsalvens en particulier. Beitr. 

geol. Karte. Schweiz. 12 : 1-273, 10 pi. 
Gray, J. E. 1848. On the Arrangement of the Brachiopoda. Ann. Mag. nat. Hist., London 

(2), 2 : 435-40- 
Greppin, J. B. 1893. Etude sur les mollusques des couches coralligenes d'Oberbuchsitten. 

Abh. schweiz. paldont. Ges., Geneva, 20 : 1-109, pi- I_ 8- 

1900. Description des fossiles du Bajocien superieur des environs de Bale. Abh. schweiz 

paldont. Ges., Geneva, 27 : 127-210, pi. 13-19, 1 table. 

Haas, .H. J. 1885-91. Etude Monographique et critique des Brachiopodes Rhetiens et 
Jurassiques des Alpes Vaudoises et des contrees environnantes. Abh. schweiz paldont. Ges., 
Geneva, 11, 14, 18 : 1-159, 21 pi., 1 table. 

1889-90. Kritische Beitrage zur Kenntniss der Jurassischen Brachiopoden fauna des 

Schweizerischen Juragebirges und seiner angrenzenden Landestheile. Abh. schweiz. 
paldont. Ges., Geneva, 16 : 1-35, pi. 1-2, 1889 ; 17 : 36-102, pi. 3-10, 1890. 



n6 JURASSIC RHYNCHONELLIDS 

Haas, H. J. & Petri, C. 1882. Die Brachiopoden der Juraformation von Elsass-Lothringen. 

Abh. geol. Spezkarte Els-Loth. 2 : i-xiv +161-320, 18 pi. Atlas. 
Hall, J. & Clarke, J. M. 1892-4. An Introduction to the Study of the Brachiopoda. N.Y. 

State Geol. Surv. (for 1891) : 135-300, pis. 1-22, 1892 ; Dept. Paleontology (for 1893) 2 : 

741-943, pi. 23-54, l8 94- 
Heer, O. 1865. Die Urwelt der Schweiz. Zurich : 1-652, pi. 1-18, figs. 1-368. 
Hendry, R. D., Rovvell, A. J. & Stanley, J. W. 1963. A rapid parallel grinding machine 

for serial sectioning of fossils. Palaeontology , London, 6, (1) : 145-7, pi. 20. 
Henry, J. 1956-8. Etude paleontologique d'Acanthotkyris spinosa Schloth., Brachiopoda 

Rhynchonellidae. Bull. Soc. linn. Normandie, Caen, (9), 9, (for 1959) : 11 9-21. 
Jacob, C. & Fallot, P. 1913. Etude sur les Rhynchonelles portlandiennes, neocomiennes et 

mesocretacees du sud-est de la France. Abh. schweiz. paldont. Ges., Geneva, 39 : 1-82, 11 pi. 
Katz, Y. I. 1962. New Genera of late Cretaceous articulate Brachiopods from the Tadjik 

Depression and the adjoining Regions. Zap. geol. Old. khar'kov. gos. Univ., 15 : 131-54, pi. 

1-2, figs. 1-16. 
Kitchin, F. L. 1900. Jurassic Fauna of Cutch. Mem. geol. Surv. India Palaeont. indica., 

Calcutta, (9), 3, (1) : 1-87, 15 pi. 
Koenig, C. D. E. 1825. Icones Fossilium Sectiles 1-4. 19 pi., 248 figs. London. 
Knorr, G. W. & Walch, J. E. I., 1768. Die Naturgeschichte der Versteinerungen, Nurnberg, 

2, (1) : 1-184, 79 pis. 
Kuhn, O. 1953. Paldogeographie des deutschen Jura, i-vi +1-74, figs. 1-34, Jena. 
Lamarck, J. B. P. 1819. Histoire naturelles des animaux sans vertebres. (1st edn.) 6, (1) : 

1-343. Paris. 
Leidhold, C. 1920. Beitrag zur genaueren Kenntnis und Systematik einiger Rhynchonelliden 

des reichslandischen Jura. Neues Jb. Geol. Paldont. (Abh. & Mh.), Stuttgart, 44 : 343-68, 

pi. 4-6. 
Lewinski, J. 1908. Les depots Jurassiques pres de la station Checiny et leur faune. Bull. 

Acad. Sci. Cracovie., 5 : 408-45, pi. 22. 
Linnaeus, C. von. 1767. Systema Naturae, 1, (2) : 533-1327, Holmiae. 
Loriol, P. de. 1896-7. Etude sur les mollusques et brachiopodes de l'Oxfordien superieur et 

moyen du Jura bernois. Abh schweiz. paldont. Ges., Geneva, 23 : 1-77, pi. 1— II, 1896 ; 24 : 

78-158, pi. 12-17, l8 77- 

- 1900. Etudes sur les mollusques et brachiopodes de l'Oxfordien inferieur ou Zone a 
Ammonites renggeri du Jura ledonien. Abh. schweiz. paldont. Ges., Geneva, 27 : 1 — 143. 6 pi. 

1904. Etude sur les mollusques et brachiopodes de l'Oxfordien superieur et moyen du Jura 

ledonien. Abh. schweiz. paldont. Ges., Geneva, (3) 31 : 161-303, pi. 20-8. 
McKerrow, W. S. 1953. Variation in the Terebratulacea of the Fuller's Earth Rock. Q. Jl 

geol. Soc. Lond., 109, (1) : 97-124, 16 figs. 
Makridin, V. P. 1955. Some Jurassic Rhynchonellids of the European part of U.S.S.R. 

Zap. geol. Fak. Kharkow State Univ., 12 : 81-91, (in Russian). 
i960. Order Rhynchonellida. In : Osnovy Paleontologii, Orlov, Yu. A. (ed.) : 239-257.pl. 

43-52, figs. 243-309. Moskva. 
1964. Brachiopods from the Jurassic of the Russian platform and some adjoining regions : 

3-339, pi. 1-25. Moscow (in Russian). 
Middlemis, F. A. 1962. Brachiopod ecology and Lower Greensand palaeogeography. 

Palaeontology, London, 5, (2) : 253-67. 
Moesch, C. 1867. Geologische Beschreibung de Aagaiier Jura Beitr. geol. Karte Schweiz. 

(4th edn.) : i-xv +1-317, 7 pi., 3 tables. 

- 1870. Paldontologische Mittheilungen aus Museum des Konigl. Bayerischen Staates. (2) 
Die Fauna der alten Cephalopoden fiihrenden Tithonbildungen. 

Moisseiev, A. S. 1939. On the stratigraphy and brachiopods of the Lower Cretaceous 
deposits of the Gagry region. (Abkhazia). Ann. Univ. Leningrad Series Geol. V7 : 
186-208, 2 pi., 2 figs. Leningrad. 



FROM NORTHWESTERN EUROPE 117 

Moisseiev, A. S. 1956. Materials on Palaeontology. New families and genera. Ministerstva 

Geologii i Okhranui. Nedr. S.S.S.R. (New ser.) 12 : 61-64. 
Munster, G. 1826. Fossiles d'Allemagne. Bull. Sci. nat. Giol., 9 : 275-72. 

1839. Beitrdge zur Petrefactenkunde. 1 : 1-124, pi. 1-18. Beyreuth. 

Muir-Wood, H. M. 1934. On the Internal Structure of some Mesozoic Brachiopoda. Philos. 

Trans., London, (B) 223 : 511-67, pi. 62-3. 

1935. Jurassic Brachiopoda. In: The Mesozoic Palaeontology of British Somaliland. 

(2) : 75-147, pi. 8-13. London. 

1936. The Brachiopoda of the British Great Oolite Series. Pt. 1. Fuller's Earth. 

Palaeontogr. Soc. [Monogr.], London : 1-144, pi. 1-5. 

1952. Some Jurassic Brachiopoda from the Lincolnshire Limestone and Upper Estuarine 

Series of Rutland and Lincolnshire. Proc. geol. Ass. Lond., 63, (2) : 13-42, pi. 5-6. 
Noetling, F. W. 1887. Der Jura am Hermon : i-v +1-46, 7 pi. Stuttgart. 
Ogerien, F. 1865. Histoire Naturelle du Jura et des Departements Voisins. 1 : i-xxvii + 1- 

947- 533 fi gs. Paris. 
Ooster, W. A. 1863. Petrifactions Remarquables des Alpes Suisses : 1— 71, 20 pi. Geneva. 
Oppel, A. 1856-8. Die Juraformation Englands, Frankreichs und des siidwestlichen Deutsch- 

lands. Wi'irtt. naturwiss. Jahreshefte. 7 : 121-558, 1856 ; 8 : 141-396, 1857 ; 9 : 129-291, 

1858. 

1865. Paldontologische Mittheilungen aus den Museum des Konigl. Bayerischen Staates. 

(5) Geognostische Studien im Ardeche Departement. 305-22. Stuttgart. 

Oppel, A. & Waagen, W. 1866. Ueber die Zone des Ammonites transver sarins. Geogn.- 

palaont. Beitr., 2 : 205-318, 1 fig. 
Orbigny, A. d'. 1847. PaUontologie francaise. Terrains Critacis (for 1848-51), 4 : 1-390, pi. 

490-599, Paris. 

1850. Prodrome de PaUontologie stratigraphique universelle. 1, (for 1849) : 1-394, Paris. 

1850. Memoire sur les brachiopodes II. Annls. Sci. nat., Paris, (3), 13 : 295-353. 

Orton, H. H. 1914. On Ciliary mechanisms in Brachiopods and some Polychaetes. /. mar. 

biol. Ass. U.K., Plymouth, 10, (2) : 283-311. 
Paetel, F. 1875. Die bisher veroffentlichen Familien — und Gattungsnamen der Mollusken : 

1-229, Berlin. 
Patrulius, D. & Orghidan, T. 1964. Contributions a l'etude de la faune Oxfordienne du 

bassin de Valea Casimcea (Dobrogea centrale). Lucr. Inst, speol. "Emil Racovita" . 3 : 

229-92, 9 pi. (in Rumanian with French summary). 
Peake, N. & B. Hancock, J. M. 1961. The Upper Cretaceous of Norfolk. Trans. Norfolk 

Norwich. Nat. Soc, Norwich, 19 (6) : 293-339, 1 pi. 9 figs., table 3. 
Pettitt, N. E. 1954. Monograph on some Rhynchonellidae of the British Chalk. Palaeon- 
togr. Soc. [Monogr.], London, (2) : 27-52, pi. III. 
Phillips, J. 1829. Illustrations of the Geology of Yorkshire ; or a Description of the Strata and 

Organic Remains of the Yorkshire Coast : 1-129, pi. 1-14. York. 

1855. The Neighbourhood of Oxford and its Geology. Oxford Essays for 1855. Oxford. 

Quenstedt, F. A. 1851-52. Handbuch der Petrefaktenkunde. (ist edn) : 1-528, pi. 1-52, 

1 85 1 : 529-792, pi. 53-62, 1852. Tubingen. 

1856-8. Der Jura : 1-576, pi. 1-72, 1856 : 577-823, pi. 73 — 100, 1857 : 824-42, 1858. 

Tubingen. 

1868-71. Petrefaktenkunde Deutschlands . II Brachiopodes 1-160, 1868 : 161-464, 1869 : 

465-748, 1870, pi. 37-61, 1871. Leipzig. 

1885. Handbuch der Petrefaktenkunde. (3rd edn) : 1-1239, 100 pi. Tubingen. 

Richardson, L. & Walker, J. F. 1907. Remarks on the Brachiopoda from the Fuller's 

Earth. Q. J I geol. Soc. Lond., 63 : 426-36, pi. 28-9. 
Roemer, F. A. 1870. Geologie von Oberschliesen : i-xxii +1-587, 50 pi. Breslau. 



n8 JURASSIC RHYNCHONELLIDS 

Rollier, L. 1917. Synopsis des Spirobranches (Brachiopodes) Jurassiques Celto-Souabes. 

Pt. II. Rhynchonellides. Abh. schweiz. palaont. Ges., Geneva, 42 : 71-184. 
Rothpletz, A. 1886. Geologisch-palaeontologische. Monographie der Vilser Alpen. Palae- 

ontographica., Stuttgart, 33 : 1-180, pi. 1-17. 
Rouillier, C. 1846. Explication de la coup geologique des environs de Moscou. Byull. 

mosk. Obshch. Ispyt. Prir., Moscow, 19, (1) : 444-85 ; (2) : 359-467, pi. A-E. 
Rousselle, L. 1965. Sur la mise en evidence par sections tranversales, du septalium des 

Rhynchonellidae (Brachiopodes). C.R. Soc. G60I. Fr., Paris, 6 : 1-207. 
Rousselle, L. & Demarez, J. P. 1964. Discrimination de deux formes infraspecifiques chez 

une Rhynchonelle du Bathonien inferieur de Bou-Rached (N.E. du Moyen-Atlas marocain). 

Bull. Soc. giol. Fr., Paris, (7), 6, (1) : 36-42, pi. iv., 1 fig. 
Rozycki, S. Z. 1948. Remarks about Upper Jurassic Rhynchonellidae of the Cracow- 

Czestochowa chain. Biul. Inst, giol., Warsaw, 42 : 16-40. (In Polish, English transla- 
tion). 
Rudwick, M. J. S. 1965. Sensory spines in the Jurassic Brachiopod "Acanthothiris ". Palae- 
ontology, London, 8, (4) : 604-17, pi. 84-7. 
Schlotheim, E. F. von. 1813. Beitrage zur Naturgeschichte der Versteinerungen in geog- 

nostischer Hinsicht. Leonard's Taschenb. fur Min. 7 : 3-134, pi. 1-4. 
1820. Die Petrefaktenkunde auf ihrem jetzigen Stand punkte durch die Beschreibung seiner 

Sammlung versteinerter und fossiler Uberberreste des Thier- und Pflanzenreichs der Vorwelt : 

i-lxii +1-437, 54 pi. Gotha. 
Schuchert, C. & Levene, C. M. 1929. Brachiopoda. Fossilium Catalogus. i. Animalia. 

(42) : 1-140, Berlin. 
Seifert, I. 1963. Die Brachiopoden des oberen Dogger der Schwabischen Alb. Palaeonto- 

graphica, Stuttgart, 121, (A) : 157-203, p. 1 10-13. 
Sherborn, D. C. 1922-33. Index Animalium. Sectio secundo Pts. 1-33. London. 
Simonescu, I. 1910. Studii Geologice si Paleontologice din Dobrogea. II. Lamelibranchi- 

atele, Gasteropodele, Brachiopodele si Echinodermili den Paturili Jurasice dela Harsova. 

Publnile Fond. Vasilie Adamachi. 25 : 1-109. 7 pi., 17 figs. 
Smirnova, T. N. 1964. K. revisi nekototorych rodov niznemelovych rinchonelid. Byull. 

mosk. Obshch. Ispyt. Prir, 135, (6) : 142-3. (in Russian). 
Smith, W. 1817. Stratigraphical System of Organized Fossils : 1-32, pi. 1-18. London. 
Sowerby, J. 1812-22. The Mineral Conchology of Great Britain. 1 : 1-234, 1812-15 ; 2 : 

1-251, 1815-18 ; 3 : 1-184, 1818-21 ; 4 (pars) : 1-113, 1821-2. London. 
Sowerby, J. de C. 1823-45. Ibid. 4 (pars) : 114-60, 1823 ; 5 : 1-116, 1823-5 ! 6 : 1-236, 

1826-9 i Index 237-250, 1835 ; 7 : 1-80, 1840-5, London. 
Struckmann, C. 1878. Der Obere fura der Umgegend von Hannover : i-viii +1-169, 8 pi. 

Hannover. 
Suess, E. 1858-9. Beitrage zur Palaeontologie von Oestereich. Die Brachiopoden des 

Stramberger Schichten. Hauer Beitr. Palaont. Ost., 1 : 15-32, pi. 1-2, 1858 ; 2 : 33-58, 

pi. 3-6. 1859. 
Szajnocha, L. 1879. Die Brachiopoden — Fauna der Oolithe von Balinbei Krakeu. Denkschr. 

Akad. Wiss. Wien, 41 : 197-240. pi. 1-7, 1 table. 
Thevenin, A. 1910. Types du Prodrome de Paleontologie stratigraphique universelle de 

d'Orbigny. Ann. PaUont., Paris, 5, (2) : 93-116, pi. 20-21. 
Thomson, J. A. 1927. Brachiopod Morphology and Genera (Recent and Tertiary). N.Z. 

Board Sci. Art Man., Wellington, 7 : 1-338, pi. 1-2, figs. 1-103. 
Thirria, M. E. 1833. Statislique Mindralogique et Giologique du Department de la Haute 

Saone. 1-465, 1 map, 1 table. Besancon. 
Thurmann, J. & Etallon, A. 1861. Lelhaea Brunlrutana, ou Eludes paUontologique et strati- 
graphique sur le Jura Bernois et en particulier les environs de Porrentruy : 1-500, 72 pi., 2 

figs., 1 map. Zurich. 



FROM NORTHWESTERN EUROPE 119 

Townsend, J. 1 81 3-1 5. The Character of Moses as established for veracity as an historian, 
recording events from the Creation to the Deluge. 2 vols. Bath. Subs. publ. (1824) in : Geo- 
logical and Mineralogical Researches during a period of more than fifty years in England, Scot- 
land, Ireland, Switzerland, Holland, France, Flanders and Spain : i-vi +1-448, pi. 1— 21. 
Bath. 

Walcott, J . 1 779- Descriptions and figures of Petrifactions found in the quarries, gravel pits, etc. 
near Bath : 1-51, 16 pi. Bath. 

Walker, J. F. 1869. On the Discovery of R. spinosa in the Bradford Clay. Proc. Yorks. nat. 
Club (for 1869) : 214-15. 

Wisniewska, M. 1932. Les Rhynchonellides du Jurassique superieur de Pologne. Palaeont. 
pol., Warsaw, 2 : 1-71, pi. 1-6. 

Weir, J. 1925. Brachiopoda, Lamellibranchiata, Gasteropoda and Belemnites. Monogr. 
geol. Dep. Hunter. Mus. (Glasg. Univ.), Glasgow, 1 : (6), 79-110, pi. 11-14. 

1929. Jurassic fossils from Jubaland, East Africa, collected by V. G. Glenday, and the 

Jurassic Geology of Somaliland. Monogr. geol. Dep. Hunter. Mus. (Glasg. Univ.), Glasgow, 
3 : 1-63, 5 pi. 

Zeuschner, L. 1845. Paleontologia Polska. Opis Zoologiczny, Botaniczny i Geologiczny . 28 
pp., 6 pi. Warsaw. 

1855. Beschreibung einer Neuen " Rhynchonella " gennant " Rhynchonella pachytheca " . 

Sitz. k. Akad. Wiss. Wien., 18, (48) : 48-50, 2 pi. 

Zieten, C. H. von. 1830-33. Die Versteinerungen Wurttembergs Verlagder expedition des 
werkes unsere Zeit : i-vi +1-102, 72 pi. Stuttgart. 



Alan Childs, B.Sc, D.I.C., Ph.D., F.G.S. 

College of Technology, 

Oxford 



PLATE i 

All specimens are figured natural size. 

In all cases: a = dorsal view. b = lateral view. c = anterior view. d = posterior view. 

Figs, i a-d. Monticlarella striocincta (Quenstedt). Kimmeridgian— cymodoce zone. 
(Bedded Virieu Limestone), N. of Lac d'Armaille (Ain), France. Collection of the Geology Dept. 
of Imperial College, London. 

Figs. 2 a-d. M. striocincta (Quenstedt). Oxfordian ("Malm alpha"). Lochengrundle, 
Germany. Collection of the University of Tubingen. 

Figs. 3 a-d. M. striocincta (Quenstedt). Details as fig. 2. 

Figs. 4 a-d. Monticlarella czenstochaviensis (Roemer). Oxfordian. Holy Cross 
Mountains, Poland. Collection of Dr. D. V. Ager. 

Figs. 5 a-d. Monticlarella triloboides (Quenstedt). Kimmeridgian, (" Weisser Jura 
delta "). " Schwabische Alb ". Collection of the University of Tubingen. 

Figs. 6 a-d, 7 a-d. M. triloboides (Quenstedt). Details as for fig. 5. 

Figs. 8 a-d. Monticlarella strioplicata (Quenstedt). Kimmeridgian, (" Malm delta "). 
Barrenberg, Germany. B.M. BB. 44130. 

Figs. 9 a-d. M. strioplicata (Quenstedt). Details as fig. 8. B.M. BB. 44131. 

Figs. 10 a-d. M. strioplicata (Quenstedt). Details as fig. 8. B.M. BB. 44132. 

Figs, ii a-d. Lacunosella arolica (Oppel). Oxfordian, transversarium zone. Jura 
meridionale, France. B.M. BB. 44134. 

Figs. 12 a-d. L. arolica (Oppel). Oxfordian, (" Birmensdorfer Schichten "), Birmensdorf, 
Switzerland. B.M. BB. 44133. 

Figs. 13 a-d. Lacunosella pseudodecorata (Rollier). ?" Weis Jura gamma ". ?Tier- 
ingen. Collection of the University of Tubingen. 



Bull. Br. Mus. nat. Hist. (Geol.) Suppl. 6 




■,3L 
la ^2a 



3a 




4a 




PLATE i 



L W 7a 



lb 



^2. #3b ^ 



4b 



5b 



6b 



I 



7b 






Ic 



^ 2c ^ 



3c 



5c 



Id 



m u *, d &„ ;d H 






6d *"7d 




PLATE 2 

Figs, i a-d. Lacunosella vaga sp. nov. Holotype. Kimmeridgian, mutabilis zone. 
Sermdrieu (Isere), France. B.M. BB. 44135. 

Fig. 2. L. vaga sp. nov. Paratype. Details as fig. 1. B.M. BB. 44136. 

Figs. 3 a-d. L. vaga sp. nov. Paratype. Kimmeridgian, (" Malm delta "). Lahm- 
Kottel, Germany. B.M. BB. 44137. 

Figs. 4 a-d. Lacunosella selliformis (Lewinski). Oxfordian. Tokarnia near Checiny, 
Poland. B.M. BB. 44142. 

Figs. 5 a-d. Lacunosella cracoviensis (Quenstedt). Upper Oxfordian-Lower Kimmer- 
idgian, (" Malm beta/gamma "). Wiirgau, Germany. B.M. BB. 44138. 

Figs. 6 a-d. L. cracoviensis (Quenstedt). Details as fig. 5. B.M. BB. 44139. 

Figs. 7 a-d. Lacunosella sparsicosta (Quenstedt). Kimmeridgian, ("Malm gamma"). 
Barrenberg, Germany. B.M. BB. 44140. 

Figs. 8 a-d. L. sparsicosta (Quenstedt). Details as fig. 7. B.M. BB. 44141. 



Bull. Br. Mus. nat. Hist. (Geol.) Suppl. 6 



PLATE 




PLATE 3 

Figs, i a-d. Lacunosella multiplicata (Quenstedt). " Weiss Jura gamma ". Thier- 
ingen, Germany. Collection of the University of Tubingen. 

Figs. 2 a-c. Lacunosella trilobataeformis Wisniewska. " Malm alpha/beta ". Riissel- 
bach bei Grafenberg, Germany. B.M. BB. 45160. 

Fig. 3. Lacunosella trilobata (Zieten). Lower Volgian, (" Malm epsilon "). Wittis- 
lingen, Germany. B.M. BB. 44143. 

Figs. 4 a-d. L. trilobata (Zieten). " Kimmeridgien ". Steinweiler bei Nattheim, Ger- 
many. Collection of the Naturhistorisches Museum, Basel, no. L 1691. 

Figs. 5 a-c, 6 a-d. L. trilobata (Zieten). Details as fig. 3. B.M. BB. 44144-5. 

Figs. 7 a-d. " Rhynchonella " pyrenaei sp. nov. Holotype. Oxfordian. Pic du Gar 
near Pau (Basses Pyrenees), France. B.M. BB. 45161. 



Bull. Br. Mus. nat. Hist. (Geol.) Suppl. 6 



PLATE 3 




7a 



&<^ 



6c 



7d 



PLATE 4 

Figs, i a-d. Rhynchonella portlandica Blake. Paratype. Lower Volgian, (Lower 
Portland Sand). St. Alban's Head, Dorset, England. B.M. B. 52284. 

Figs. 2 a-d. R. portlandica Blake. Holotype. Details as fig. 1. B.M. B. 52283. 

Figs. 3 a-d. Rhynchonella loxiae Fischer. Upper Volgian. Khoroshevo near Moscow. 
Collection of Dr. D. V. Ager. 

Figs. 4 a-b. Rhynchonella cf. R. portlandica Blake. Lower Volgian, (Portland Sands). 
Ringstead Bay, Dorset, England. B.M. B. 72240. 

Figs. 5 a-d. Rhynchonella cf . R. portlandica Blake. Lower Volgian, (Basal Shell Bed, 
Portland Limestone). Cliff N.W. of Portland, England. Collection of the University Museum, 
Oxford. J. 1006. 

Figs. 6 a-d, 7, 8. Rhynchonella sp. Kimmeridgian, Ipseudomutabilis zone, (Boulder Beds). 
West Garty, Sutherland, Scotland. B.M. BB. 44146-7, 45162. 

Figs. 9 a-c, 10 a-d, 11 a-d, 12 a-d. Rhynchonella cf. R. portlandica Blake. Lower 
Volgian, (Upper Blacknore Beds). West Weare Cliffs, Portland, England. Collection of the 
University Museum, Oxford. J. 2462. 

Figs. 13 a-d, 14 a-d. Rhynchonella cf. R. portlandica Blake. Lower Volgian, (upper part 
of Portland Sand). Foot of Gad Cliff, Dorset, England. Collection of the University Museum, 
Oxford. J. 167. 

Figs. 15 a-d. Rhynchonella cf. R. subvariabilis Davidson. Lower Volgian, (top of Kim- 
meridge Clay just below Portland Sands Passage Beds). Chapman's Pool, Dorset, England. 
B.M. B. 47251. 

Figs. 16 a-d. Rhynchonella cf. R. subvariabilis Davidson. Lower Volgian, (Kimmeridge 
Clay, about 40' below rotundum nodules). Chapman's Pool, Dorset, England. Collection of the 
Sedgewick Museum, Cambridge. J. 1500. 

Figs. 17 a-d. Rhynchonella cf. R. sarapaulensis Makridin. Lower Volgian, (Kimmeridge 
Clay, rotundum nodules). Chapman's Pool, Dorset, England. Collection of the Sedgewick 
Museum, Cambridge. J. 1512. 

Figs. 18 a-d. Rhynchonella cf. R. portlandica Blake. Lower Volgian, (Lyddite Bed, 
Portland Beds). Long Crendon, Buckinghamshire, England. Collection of the Sedgewick 
Museum, Cambridge. J. 1524. 



Bull. Br. Mus. nat. Hist. (Geol.) Suppl 6 

O h <fr 



PLATE 4 






lb 2a 

/J '^ 



lc Id 






2c 2d 







3b 





£ i 





5b 



5c >**<|B r 5d 



v. 




6a 




6b 



^mbJBP 6d 4^4X^P 




3c ^^ 4b 



Ai 



6c 






h 




9a 9b 



9c 



l2b 'ftWlla 



Mb 







10b 



10a 







1 2d 



3a 



</a 





12c 



13b 




lc 




10c |0d 



lid 




13c 





13d 





/. 



w,» 4 





1 6a 



dSBi 




1 5c 





14b 



I4d 




4a 







14c 



16b 




16c 




17c 





18c 




I5d 




7b 



s* 



I7d 




8a 





18b 



I8d 



PLATE 5 

Figs, i a-d, 2 a-d. Rhynchonella sp. Callovian. Wrzova, Poland. B.M. BB. 44148-9. 

Figs. 3 a-d, 4 a-d, 5, 6. Rhynchonella rivelensis Loriol. Upper Oxfordian. Alex near 
Oyonnax (Ain), France. B.M. BB. 44150-53. 

Figs. 7 a-d, 8 a-d, 9 a-c. Thurmannella obtrita (Defrance). Oxfordian, (Terrain a 
Chailles). Pagny-sur-Meurse (Meurthe et Moselle), France. B.M. BB. 44154-56. Figs. 8 
a-d neotype. 

Figs. 10 a-d. Thurmanella acuticosta sp. nov. Oxfordian, (Lower Calcareous Grit). 
Filey Brigg, Yorkshire, England. Holotype. B.M. BB. 44158. 

Figs, ii a-d. T. acuticosta sp. nov. Paratype. Details as fig. 10. B.M. BB. 44157. 

Figs. 12 a-d. Acanthothiris spinosa (Linnaeus). Neotype. Bajocian, (Upper Trigonia 
or Clypeus Grit) . Rodborough Hill, Gloucs., England. B.M. BB. 45400. 

Figs. 13 a-d. A. spinosa (Linnaeus). Bajocian, (Upper Trigonia or Clypeus Grit). Stroud, 
Gloucs., England. B.M. B. 63318. 

Figs. 14 a-d. A. spinosa (Linnaeus). Juvenile specimen. Bajocian, (Upper Trigonia 
Grit). Waterworks Quarry, Wotoon under Edge, Gloucs., England. B.M. BB. 44162. 

Fig. 15. A. spinosa (Linnaeus). Crushed specimen showing spines attached. Upper 
Bathonian, (" Varians Schichten "). Liestal near Basel, Switzerland. B.M. BB. 44160. 

Figs. 16 a-d, 17 a-d. A. spinosa (Linnaeus). Details as fig. 15. B.M. BB. 44161 and 44159. 



Bull. Br. Mus. nat. Hist. (Geol.) Suppl. 6 



PLATE 5 



d M <^ 




PLATE 6 

Figs, i a-d. Plaster cast of the type specimen Orbigny's Acanthothiris costata the original is 
in the collection of Jardin des Plantes (Musee d'histoire Naturelle), Paris. This specimen was 
figured by Thevenin (1910). 

Figs. 2 a-d, 3 a-d, 4 a-d. Acanthothiris cf. A. costata Orbigny. Upper Bajocian-Lower 
Callovian (Chanaz Beds). Pugieu (Ain), France. B.M. BB. 44163-65. 

Figs. 5 a-d. Acanthorhynchia (Acanthorhynchia) panacanathina (Buckman & 
Walker). Bajocian, zigzag zone. Burton Bradstock, Dorset, England. B.M. B. 63479. 

Figs. 6 a-d. A. (Acanthorhynchia) panacanathina (Buckman & Walker). " Inferior 
Oolite ". Larkfield Quarry, Burton Bradstock, Dorset, England. B.M. B. 31306. 

Figs. 7 a-d. A. (Acanthorhynchia) panacanthina (Buckman & Walker). Bajocian, 
parkinsoni zone. Wennet Hills, Burton Bradstock, Dorset, England. B.M. B. 31304. 

Figs. 8 a-d. Acanthorhynchia (Acanthorhynchia) spinulosa (Oppel). " Oxfordien ". 
Montsaon, Chaumont (Haute Marne), France. B.M. B. 36080. 

Figs. 9 a-d. A. (Acanthorhynchia) spinulosa (Oppel). " Oxfordien ". Bolognes 
(Haute Marne), France. B.M. B. 36079. 

Figs. 10 a-d. Acanthorhynchia (Acanthorhynchia) vilsensis sp. nov. Paratype. 
Callovian. Legam bei Vils, southern Germany. Collection of the Bayerische Staatsammlung 
fur Palaontologie und historische Geologie, Munich. AS xxiv 5. 

Figs, ii a-d. A. (Acanthorhynchia) vilsensis sp. nov. Holotype. Details as fig. 10. 
AS xxiv 2. 

Figs. 12 a-d. A. (Acanthorhynchia) vilsensis sp. nov. Paratype. Details as fig. 10. 
AS xxiv 5. 



Bull. Br. Mus. nat. Hist. (Geol.) Suppl. 6 



PLATE 6 




PLATE 7 

Figs, i a-d. Acanthorhynchia (Echinirhynchia) senticosa (Schlotheim). Neotype. 
" Weissem Jura ". Amberg, Germany. Figured by Rothpletz (1886). Collection of the 
Bayerische Staatsammlung fur Palaontologie und historische Geologie, Munich. AS vii 330. 

Figs. 2 a-d, 3 a-d. A. (Echinirhynchia) senticosa (Schlotheim). "Malm delta". 
Heiligenstadt near Erlangen, Germany. Collection of the Geologisches Institut der Universitat, 
Erlangen. 

Figs. 4 a-d, 5, 6. Acanthorhynchia (Echinirhynchia) fileyensis (Buckman & Walker). 
Oxfordian, (Lower Calcareous Grit). Filey, Yorkshire, England. B.M. B. 31867, B. 31334, B. 

3I333- 

Figs. 7 a-d. Acanthorhynchia (Echinirhynchia) lorioli (Rollier). Oxfordian, (Terrain 
a Chailles). Fringeli, " Berner Jura ". Collection of the Naturhistorisches Museum, Basel. 
L 1700. 

Figs. 8 a-d, 9. A. (Echinirhynchia) lorioli (Rollier). Oxfordian, transverarium zone. St. 
Sorlin, Liefnans (Jura), France. B.M. BB. 44166-67. 

Figs. 10 a-b, 11 a-d. Acanthorhynchia sp. Upper Kimmeridgian, (Oignon Limestone). 
Colomieu (Ain), France. B.M. BB. 44169, 44168. 

Figs. 12 a-d, 13, 14, 15, 16. Somalirhynchia sutherlandi (Davidson). Topotypes. 
Kimmeridgian, Ipseudomutabilis zone (Boulder Beds). West Garty, Sutherland, Scotland. 
Collection of the Geological Survey of Great Britain (Edinburgh Office). Fig. 12 no. G.S.E. 3684, 
West Garty; fig. 13 no. M 2268g, West Garty ; fig. 14 no. G.S.E. 3681, West Garty ; fig. 15 no. 
G.S.E. 3685, West Garty ; fig. 16 no. G.S.E. 3686, Portgower. 



Bull. Br. Mus. nat. Hist. (Geol.) Suppl. 6 

fa 



PLATE 7 




PLATE 8 

Figs, i a-d, 2 a-d, 3. Somalirhynchia moeschi (Haas). Oxfordian, pseuaocordata zone. 
Rochers de Fresnois, Chaumont, pres Saint-Claude (Jura), France. B.M. BB. 44170-72. 

Figs. 4 a-d, 5 a-d. Septaliphoria arduennensis (Oppel). Oxfordian, (Terrain a Chailles). 
Pagny-sur-Meurse (Meurthe et Moselle), France. Fig. 4 neotype. B.M. BB. 44173-74. 

Figs. 6 a-d, 7 a-d. Septaliphoria? septentrionalis sp. nov. PUpper Oxfordian-Plower 
Kimmeridgian, (Alt-na-Cuile Sandstone). Alt-na-Cuile, Sutherland, Scotland. Fig. 6 holotype ; 
fig. 7 paratype. B.M. BB. 44175-76. 



Bull. Br. Mus. nat. Hist. (Geol.) Suppl. 6 



PLATE 8 




PLATE 9 

Figs, i a-d, 2 a-c, 3 a-c, 4. Septaliphoria? hudlestoni (Rollier). Topotypes. Kimmer- 
idgian, (Abbotsbury Ironstone). Abbotsbury, Dorset, England. B.M. 

Figs. 5 a-c, 6 a-d. Septaliphoria paucicosta sp. nov. Oxfordian, (Lower Calcareous 
Grit). Hutton Bushell, Beedale, Yorkshire. 

Fig. 5. Paratype ; BB. 45395, fig. 6 holotype. B.M. BB. 45394. 

Figs. 7 a-d, 8 a-d, 10, 11. " Rhynchonella " ordinaria sp. nov. Paratypes. " Sequan- 
ien ". Calliors du Chateau, Bourges (Cher), France. B.M. BB. 45163-66. 

Figs. 9 a-d. " Rhynchonella " ordinaria sp. nov. Holotype. Oxfordian, pseudocordata 
zone. Mont Dolet, Serm6rieu (Isere), France. B.M. BB. 45167. 



Bull. Br. Mus. nat. Hist. (Geol.) Suppl. 6 



PLATE 9 




PLATE 10 

Figs, i a-d, 2 a-d, 3 a-d. Torquirhynchia inconstans (J. Sowerby). Topotypes. 
Kimmeridgian, baylei zone, (Kimmeridge Clay). Ringstead Bay, Dorset, England. B.M. BB. 
44177-79. 

Figs. 4 a-c, 5 a-d, 6. Torquirhynchia guebhardi (Jacob & Fallot). Kimmeridgian, 
cymodoce zone, (Bedded Virieu Limestone). N. of Lac d'Armaille (Ain), France. B.M. BB. 
45168-70. 



Bull. Br. Mus. nat. Hist. (Geol.) Suppl. 6 



PLATE 10 




PLATE ii 



Figs, i a-d, 2 a-d, 3 a-d. Torquirhynchia speciosa (Miinster). Upper Kimmeridgian- 
Lower Volgian, (" Diceras-Kalk "). Saal near Ingolstadt, Germany. B.M. BB. 45171-73. 



Bull. Br. Mus. nat. Hist. (Geol.) Suppl. 6 



PLATE ii 




PLATE 12 

Figs, i a-d, 2 a-d, 3 a-d. Torquirhynchia cf. T. astieriformis (Wisniewska). Kimmeridg- 
ian, mutabilis zone. Sermerieu (Isere), France. B.M. BB. 45174-6. 

Fig. 5. Torquirhynchia inconstans (Sowerby) from the Kimmeridge Clay of Ringstead 
Bay, Dorset. Shows the nature of the articulation, the shape of the median ridge and the out- 
line of the muscle scars in the brachial valve. x 1%. B. 60368. 

Fig. 4. Transverse through the hinge plate of Septaliphoria paucicosta sp. nov. showing 
the form of the septalial plate and the crural base. This is part of section 4, 3 figured in text-fig. 
33- X12. 

Fig. 6. Transverse section of Septaliphoria paucicosta sp. nov. showing the median 
septum and septalial plates. This is part of section 3, 4 in the series shown in text-fig. 33. x 5. 

Fig. 7. Transverse section of Acanthorhynchia (Acanthorhynchia) panacanthina 
(Buckman & Walker) showing the thickened nature of the hinge plates. This is part of section 
i, 9 in the series shown in text-fig. 24. 

Fig. 8. Transverse section of Zeilleria austriaca (Zugmeyer) showing the form of the 
septalium. Specimen obtained from the Rhaetian of Piesting Tal, Lower Austria. The " peel " 
from which the photomicrograph was taken was kindly made available to the author by Dr. 
D. A. B. Pearson. X5. 

Fig. 9. Transverse through the hinge plates of Lacunosella vaga sp. nov. showing the form 
of the falcifer crura. This is part of section 3, 6 in the series shown in text-fig. 14. 



Bull. Br. Mus. nat. Hist. (Geol.) Suppl. 6 



PLATE 12 







LIST OF SUPPLEMENTS 
TO THE GEOLOGICAL SERIES 
OF THE BULLETIN OF 
THE BRITISH MUSEUM (NATURAL HISTORY) 



i. Cox, L. R. Jurassic Bivalvia and Gastropoda from Tanganyika and Kenya. 
Pp. 213; 30 plates, 2 Text-figures. 1965. £6. 

2. El-Naggar, Z. R. Stratigraphy and Planktonic Foraminifera of the Upper 
Cretaceous — Lower Tertiary Succession in the Esna-Idfu Region, Nile Valley, 
Egypt, U.A.R. Pp. 291 ; 23 plates, 18 Text-figures. 1966. £10. 

3. Davey, R. J., Downie, C, Sargeant, W. A. S. & Williams, G. L. Studies on 
Mesozoic and Cainozoic Dinoflagellate Cysts. Pp. 248; 28 plates, 64 Text- 
figures. 1966. £y. 

4. Elliott, G. F. Permian to Palaeocene Calcareous Algae (Dasycladaceae) of the 
Middle East. Pp. in; 24 plates, 17 Text-figures. 1968. £5 2s. 6d. 

5. Rhodes, F. H. T., Austin, R. L. & Druce, E. C. British Avonian (Carboni- 
ferous) Conodont faunas, and their value in local and continental correlation. 
Pp- 3 1 5> 3 1 plates, 92 Text-figures. 1969. £11. 



Primed in England by Staples Printers Limited at their Kettering, Nortbants, establishment 




THE RELATIONSHIPS OF CERTAIN 

UPPER CRETACEOUS TELEOSTS 

WITH SPECIAL REFERENCE TO THE 

MYCTOPHOIDS 



P. C. GOODY 



BULLETIN OF 
THE BRITISH MUSEUM (NATURAL HISTORY) 
GEOLOGY Supplement 7 

LONDON: 1969 



THE RELATIONSHIPS OF CERTAIN UPPER 

CRETACEOUS TELEOSTS WITH SPECIAL 

REFERENCE TO THE MYCTOPHOIDS 




BY 

PETER C. GOODY 



H 



102 Text-figures 



BULLETIN OF 
THE BRITISH MUSEUM (NATURAL HISTORY) 
GEOLOGY Supplement 7 

LONDON: 1969 



THE BULLETIN OF THE BRITISH MUSEUM 

(natural history), instituted in 1949, is 
issued in five series corresponding to the Departments 
of the Museum, and an Historical series. 

Parts will appear at irregular intervals as they become 
ready. Volumes will contain about three or four 
hundred pages, and will not necessarily be completed 
within one calendar year. 

In 1965 a separate supplementary series of longer 
papers was instituted, numbered serially for each 
Department. 

This paper is Supplement 7 of the Geological 
(Palaeontological) series. The abbreviated titles of 
periodicals cited follow those of the World List of 
Scientific Periodicals. 



World List abbreviation 
Bull. Br. Mus. not. Hist. (Geol.) Suppl. 



Trustees of the British Museum (Natural History) 1969 



TRUSTEES OF 
THE BRITISH MUSEUM (NATURAL HISTORY) 

Issued 3 October, 1969 Price £6 10s, 



THE RELATIONSHIPS OF CERTAIN UPPER 

CRETACEOUS TELEOSTS WITH SPECIAL 

REFERENCE TO THE MYCTOPHOIDS 

By PETER C. GOODY 



CONTENTS 



I. Introduction .... 
II. Systematic descriptions . 

Superorder Protacanthopterygii 
Order Salmoniformes 

Suborder Ichthyotringoidei 

Family Ichthyotringidae Jordan 
Genus Ichthyotringa Cope . 

Ichthyotringa fuvcata (Agassiz) 
Ichthyotringa delicata (Hay) 
Incertae sedis : Genus Apateodus Woodward 

Apateodus stviatus Woodward 
Family Apateopholidae nov. 
Genus Apateopholis Woodward . 

Apateopholis laniatus (Davis) 
Suborder Cimolichthyoidei 
Family Cimolichthyidae nov. 
Genus Cimolichthys Leidy . 

Cimolichthys levesiensis Leidy 
Family Dercetidae Pictet . 
Genus Dercetis Agassiz 

Dercetis triqueter Pictet 
Genus Rhynchodercetis Arambourg 

Rhynchodercetis hakelensis Pictet & Humbert 
Rhynchodercetis gortanii (d'Erasmo) 
Genus Pelargorhynchns von der Marck . 

Pelargorhynchus dercetiformis von der Marck 
Suborder Enchodontoidei 

Family Enchodontidae Woodward 
Genus Enchodus Agassiz 

Enchodus lewesiensis (Mantell) . 
Enchodus marchesettii (Kramberger) 
Genus Palaeolycus von der Marck 

Palaeolycus dreginensis von der Marck 
Family Eurypholidae nov. . 
Genus Eurypholis Pictet 

Eurypholis boissieri Pictet 
Eurypholis pulchellus (Woodward) 
Genus Saurorhamphus Heckel 

Saurorhamphus freyeri Heckel . 
Suborder Halecoidei 

Family Halecidae Agassiz 
Genus Halec Agassiz . 

Halec eupterygius (Dixon) . 
Halec haueri (Bassani) 
Genus Phylactocephalus Davis 



Page 

4 

6 

6 

6 

6 

7 

7 

7 

14 

17 

18 

30 
30 
30 
36 
36 
37 
37 
50 
5i 
52 
59 
59 
65 
67 
68 
7i 
7i 
7i 
72 

85 

95 

95 

99 

99 

99 

no 

121 

122 

126 

126 

126 

127 

137 
140 



UPPER CRETACEOUS TELEOSTS 



Phylactocephalns microlepis Davis 
Genus Hemisaurida Kner . 

Hemisaurida hakelensis sp. nov. 
Order Myctophiformes 

Family Sardinioididae nov. 

Genus Sardinioides von der Marck 

Sardinioides minimus (Agassiz^ 
Order Ctenothrissiformes. 

Family Aulolepidae Patterson 
Genus P alter sonichthys gen. nov. 

Patter sonichthys delicatus sp. nov, 

III. Discussion ...... 

Order Salmoniformes 

(A) Suborder Ichthyotringoidei . 

(B) Suborder Cimolichthyoidei 

Family Prionolepididae nov. 
Genus Prionolepis Egerton . 

Prionolepis cataphractus (Pictet 

(C) Suborder Enchodontoidei 

(D) Suborder Halecoidei 
Order Myctophiformes 
Order Ctenothrissiformes. 

IV. Teleost evolution within the cretaceous 
V. Conclusions ...... 

VI. Acknowledgements .... 

VII. References ...... 

VIII. Lettering used in the text-figures 



& Humbert) 



141 
146 
148 
153 
153 
153 
153 
160 
160 
160 
160 
169 
169 
169 
176 
178 
179 
179 
191 
198 
202 
229 

233 
246 
247 
248 
253 



SYNOPSIS 

Redescriptions are given of most of the genera contained in Woodward's (1901) families 
Enchodontidae, Dercetidae and Scopelidae. Four new suborders are erected within the order 
Salmoniformes, the Ichthyotringoidei, Cimolichthyoidei, Enchodontoidei and Halecoidei. The 
Ichthyotringoidei contains two families, the Ichthyotringidae and the new family Apateopho- 
lidae. The Cimolichthyoidei contains the Dercetidae and the new family Cimolichthyidae, 
with a third new family appended, the Prionolepididae. The Enchodontoidei is composed of 
two families, Enchodontidae and the new family Eurypholidae. The Halecoidei contains the 
single family Halecidae. A new family Sardinioididae is added to the Myctophiformes and a 
new genus, Patter sonichthys , is included within the ctenothrissiform family Aulolepidae. Each 
group is discussed in relation to its systematic position and the information gained is used in an 
overall consideration of the evolution of the teleosts in the Cretaceous. 



I. INTRODUCTION 

The present work is intended as a revision of certain lower teleostean fishes from the 
Upper Cretaceous. By the close of this era teleosts were numerous and varied and 
many of the present day suborders were represented, the notable exception being the 
perciform groups (see Patterson, 1967b). The faunal composition at the base of the 
Upper Cretaceous appears to have been almost as varied. Patterson (1964, 1967a) has 
dealt extensively with the Ctenothrissiformes and the acanthopterygian faunas from 



ESPECIALLY MYCTOPHOIDS 5 

the Upper Cretaceous, however teleosts below the acanthopterygian grade have re- 
ceived no modern comparable treatment, except for the work of Dunkle (1940) on the 
elopoid Notelops brama, and Bardack (1965) on the ichthyodectids. 

The important earlier works on Cretaceous fish faunas are those of Agassiz (1833- 
1844), Heckel (1849), Pictet (1850), Pictet and Humbert (1866) and Davis (1887), and 
with the exception of Agassiz were all primarily concerned with material from the 
Lebanon. Dixon (1850) described forms from the English Chalk; von der Marck 
(1858, 1863) those from the Westphalian Chalk; and both Bassani (1882) and Kram- 
berger (1895) considered the Dalmatian Chalk. The information contained in these 
and other earlier works was collected together and reinterpreted by Woodward 
(1901). The fish fauna of the English Chalk was later extensively treated in a mono- 
graph by Woodward (1902-1912). More recently d'Erasmo (1946) has worked on 
the Upper Cretaceous fauna from Comen near Trieste, Siegfried (1954) has reviewed 
von der Marck's Westphalian material, Arambourg (1954) has described a new fauna 
from Morocco, and Leonardi (1966) has reported on a Sicilian fauna. 

Originally the work was begun as a revision of Woodward's (1901) family Encho- 
dontidae. It was hoped that a study of the poorly known genera included in this 
family would yield possible ancestral forms for the present day Myctophiformes. 
This view had been expressed by Woodward (1901: ix) when he stated that " the 
Enchodontidae might perhaps furnish the ancestors" of the isospondylous scopeloids. 
All of the genera in the family have been considered and most are described at some 
length in the following pages. Only those species represented by reasonably com- 
plete material have been included, species erected on isolated teeth or jaw fragments 
are omitted. The Upper Cretaceous fauna of the United States has not been treated 
due to the absence of good material in collections outside America. 

During the study of the Enchodontidae it soon became clear that many of the 
included genera were not at all closely related. In order to elucidate the systematic 
position of some of these genera it was necessary to consider two more of Woodward's 
(1901) families, the Dercetidae and Scopelidae. Many of the genera were found to be 
not only widely separated, but referable to different suborders and even orders. 
Consequently recent representatives of the orders Elopiformes, Salmoniformes and 
Myctophiformes have also been studied. 

The first part of the work is devoted to the detailed description of the fossil genera, 
and is set out in sequence from the most primitive Salmoniformes up through the 
Myctophiformes to the Ctenothrissiformes. The findings concerned with the Elopi- 
formes have been published elsewhere (Goody, 1969). In the second part, each of 
these major fossil groups is considered in relation to its systematic position, while 
certain morphological problems and possible modes of life are also discussed. Thirdly 
the information set out in the first two parts is used in an overall consideration of 
teleost evolution within the Cretaceous. 

Most of the fossil material used is from the collections in the British Museum 
(Natural History), referred to, throughout this work, as the B.M.N.H. Type material 
has been examined wherever possible and certain specimens have been loaned from 
museums in Europe. Where such material has been examined, mention is made in 
the text. 



6 UPPER CRETACEOUS TELEOSTS 

Preparation of fossil material was mainly accomplished by dissolving the matrix in 
acetic acid (Toombs, 1948 ; Rixon, 1949). Specimens 'in the round' were immersed in 
2 % to 5% acetic acid, for periods of several hours, after which they were washed in 
running water and then thoroughly dried. The subsequently exposed bone was 
hardened by the application of a solution of Vinalak 5917 in methyl ethyl ketone. 
The Vinalak was most effective when used in low concentrations. In two cases, 
where the material had previously been figured, a replica of the original was obtained 
by making silicone rubber moulds which were employed for the production of plaster 
of paris casts. 

Acid preparation was also used in conjunction with the embedding of flattened 
specimens in a cold-curing polyester resin, Crystic 195, the transfer method of 
Toombs and Rixon (1950). Excellent preparations of specimens from the Middle 
Cenomanian deposits of Hakel and Hajula in the Lebanon were obtained by this 
method. The matrix in this case is a much harder, more fissile limestone than is the 
English Chalk and greater concentrations of acetic acid, up to 20%, were used. 

The recent material used was either skeletal or alizarin preparations. In the case 
of the Myctophiformes the specimens from which bone and alizarin preparations 
were made were kindly loaned by Dr. G. Maul of Funchal in Madeira. 

Certain teleost bones have been given varying names throughout the literature. 
The nomenclature adopted in this work agrees with that of Patterson (1964), with the 
addition that the caudal skeleton terminology follows that of Nybelin (1963). In the 
systematic descriptions the vertebral counts always exclude the ural vertebrae and 
any preural vertebrae to which these may fuse, thus the count extends to the first 
free preural vertebra. Also in the systematic descriptions caudal vertebrae are 
those vertebrae which possess a complete haemal arch. Finally the classification 
adopted is that proposed by Greenwood, et al. (1966). 



II. SYSTEMATIC DESCRIPTIONS 

Superorder PROTACANTHOPTERYGII 

Order SALMONIFORMES 

Suborder ICHTHYOTRINGOIDEI 

Diagnosis. Head elongated anteriorly into a prominent rostrum ; body short. 
Maxilla enters gape, toothed or untoothed. Teeth on premaxillae, palatines, 
ectopterygoid, endopterygoid and dentary ; vomer untoo thed. 9 branchiostegal 
rays. Pectoral fins with up to 20 rays, low on flanks ; mesocoracoid arch present ; 
pectoral girdle with 2 postcleithra. Epineurals along most of the body. Anterior 
neural spines consist of separate lateral elements. Caudal skeleton with 2 free ural 
vertebrae ; 19 principal rays of which 17 are branched. Fins without spines. 
Scales cycloid. 



ESPECIALLY MYCTOPHOIDS 7 

Family ICHTHYOTRINGIDAE Jordan, 1905 

Diagnosis. Head excessively elongated and shallow ; body unelongated. 
Post-temporal fossa roofed. Parietals meet in the mid-line. Orbitosphenoid and 
basiphenoid present. Posterior infraorbital bones expanded. Mandibular sus- 
pensorium vertical or near vertical. Preoperculum with no ventral spine. Vertebral 
elements relatively well fused ; centra longer than deep. Body covered with thin 
cycloid scales enlarged along the lateral line. Dermal bones smooth, devoid of 
ornamentation. 

Genus ICHTHYOTRINGA Cope, 1878 

Diagnosis (emended). Ichthyotringidae in which the head is very shallow and 
elongated. Dermal bones of the head smooth and unornamented. Jaw suspension 
vertical or inclined very slightly forwards. Branchiostegal rays approximately 9 in 
number. Vertebrae 40 in number, of which 17 are caudal. Individual centra longer 
than deep. Pectoral fin with up to 20 rays. Pelvic fins abdominal with as many 
as 12 rays. 

Type species. Ichthyotringa tenuirostris Cope. 

Remarks. Cope's generic name Ichthyotringa (1878:69) replaces Rhinellus 
Agassiz (1844 : 260) which is preoccupied. 

Ichthyotringa J areata (Agassiz) 
(Text-figs. 1-4) 

1844 Rhinellus furcatus Agassiz , 2 : 260, pi. 58ft, fig. 5. 



1849 Rhinellus furcatus Agassiz 

1850 Rhinellus furcatus Agassiz 
1866 Rhinellus furcatus Agassiz 
1873 Rhinellus furcatus Agassiz 
1887 Rhinellus furcatus Agassiz 
1 90 1 Rhinellus furcatus Agassiz 
1908 Rhinellus furcatus Agassiz 
1932 Rhinellus furcatus Agassiz 
1954 Rhinellus furcatus Agassiz 



Heckel : 340, pi. 23, fig. 16. 

Pictet : 44, pi. 8, figs. 3, 4. 

Pictet & Humbert : 82, pi. 11, figs. 5- 

von der Marck : 60, pi. 2, fig. 4. 

Davis : 608. 

Woodward : 266, pi. 13, figs. 1, 2. 

Priem : 10, pi. 1, fig. 13. 

Woodward : 162, fig. 249. 

Siegfried : 21. 



1964 Ichthyotringa fur cata (Agassiz) Danil'chenko : 418, fig. 140. 
1966 Rhinellus furcatus Agassiz ; Lehman : 203, fig. 201. 

Diagnosis (emended). Ichthyotringa of standard length not exceeding 10 cm. 
Head with opercular apparatus occupies half the standard length. Mandibular 
suspensorium vertical. Dorsal fin with 12 rays, mid-way between occiput and 
caudal peduncle. Anal fin with 15 rays arising behind dorsal fin, nearer to caudal 
than to pel vies. Pelvics in advance of dorsal fin with 11 rays. Vertebral column 
with 40 vertebrae. 

Holotype. Present location unknown to the author, at one time in the Amic 
collection in Paris. The holotype was from the Upper Senonian of Sahel Alma, 
Lebanon. 



8 UPPER CRETACEOUS TELEOSTS 

Material. B.M.N.H. numbers 48089a, 480896, 48092, 49544, all from the Upper 
Senonian of Sahel Alma, Lebanon, and all prepared by the transfer method in acetic 
acid. Specimen number 8436, being the negative impression of a complete fish in the 
Palaontologisches Institut der Westfalische Wilhelms-Universitat, Miinster. This 
latter specimen was that used by von der Marck (1873 : 60, pi. 2, fig. 4), and is from 
the Upper Senonian of Sendenhorst, Westphalia. 

Description. Neurocranium. A dorsal view of the reconstructed neurocranium 
is shown in Text-figure 1. The neurocranium is elongate, shallow, relatively narrow, 
and drawn out preorbitally to enter into the composition of the rostral region. The 
mesethmoid extends the whole length of the rostrum and so too does the vomer 
below it. In the base of the orbit the parasphenoid contacts the vomer anteriorly in 
front of the lateral ethmoids. The frontals form most of the skull-roof and are un- 
ornamented, but bear two large lateral ridges which run longitudinally above the 
orbital region and demarcate a median interorbital depression. Lateral to the 
ridges, above the orbit, the frontals are curved, forming the upper margin to the 
orbit. 

The supraoccipital is small and contacts the parietals anteriorly and the epiotics 
laterally. The parietals are large and meet in the mid-line of the skull-roof anterior 
to the supraoccipital. Laterally the parietals curve ventro-laterally and are over- 
lapped by the pterotics, forming, presumably, the roof of the post-temporal fossa. 




e P° soc 

Fig. 1. Ichthyotringa furcata (Agassiz). Neurocranium in dorsal view. 



ESPECIALLY MYCTOPHOIDS 9 

The sphenotic makes up the postero-dorsal corner of the orbit, contacting the 
pterotic posteriorly and the frontal dorso-medially. The pterotic forms the postero- 
lateral border of the skull-roof and is inclined ventro-laterally. The anterior region 
of the hyomandibular facet is composed of sphenotic with the possibility of a con- 
tribution from the prootic, while the main part of the facet is formed from pterotic. 
The epiotics form the postero-lateral region of the occipital border of the roof bound- 
ing the parietals posteriorly. The dorsal region of the epiotic is thickened to produce 
a knob of bone with which the dorsal limb of the post-temporal articulated. 

Within the posterior region of the orbit, traces of median bony elements are 
observable which may correspond to an orbitosphenoid and a basisphenoid, but may 
simply be other neurocranial bones which have become displaced. 

Jaws and suspensorium. The dermal upper jaw is extended into a long thin 
rostrum, with contributions from the palate and neurocranium. The hyomandibular 
has an elongated head with a lateral crest passing back to the rear edge of the bone. 
The hyomandibular is held vertically and the quadrate condyle is immediately below 
the occiput. The quadrate is a shallow, triangular bone with an expanded and 
extremely well ossified ventral condyle. The condyle is transversely orientated and 
fits into a shallow articular facet. The quadrate meets the ectopterygoid anteriorly 
and the metapterygoid dorsally. The metapterygoid is clearly visible overlapping 
the anterior regions of the lateral face of the hyomandibular. The anterior edge of 
the metapterygoid is produced into an oblique thickened ridge which is continuous 
with a ridge on the postero-dorsal part of the ectopterygoid. The ectopterygoid is 
thin and elongated, posteriorly contacting the quadrate and the metapterygoid, and 
anteriorly the palatine. The connection with the palatine is in the region of the 
lateral ethmoid at the anterior end of the orbit. The palatine continues forwards as 
a direct extension of the ectopterygoid along the entire length of the rostrum. The 
palatine bone is supported medially by the extended mesethmoid and vomer and 
takes the form of a thin vertical lamina of bone. Its ventral edge bears a single 
row of relatively large, slightly recurved needle-like teeth. 

The dermal upper jaw is composed of premaxilla and maxilla, although it is practic- 
ally impossible to determine to what extent the maxilla enters the oral border. It 
would appear that at least the anterior two-thirds of the gape is produced by the 
premaxilla and possibly more. The dermal upper jaw bears a marginal row of 
minute teeth. 

The mandible is long and shallow bearing two rows of teeth, a marginal row of 
minute teeth and an inner row of larger teeth approaching the size and form of those 
on the palatine. Almost all of the mandible is composed of dentary, the articular 
being a relatively small component. The articular facet is shallow, despite the size of 
the quadrate condyle, and there is no coronoid or retroarticular process. 

Opercular bones. The preoperculum is narrow and upright with a slight basal 
expansion. The anterior edge of the preoperculum abuts against the posterior face 
of the hyomandibular crest dorsally, and against the rear edge of the quadrate 
ventrally. The operculum is large, slightly deeper than it is long with a rounded 
upper and posterior margin and a straighter, postero-dorsally inclined ventral edge. 
The suboperculum is large with rounded ventral and posterior edges. Dorsally it is 



UPPER CRETACEOUS TELEOSTS 



attached to the internal face of the operculum. A small interoperculum appears 
to be present, below the suboperculum and near to the ventral preopercular region. 
The opercular bones are smooth, as are the other dermal skull bones. 



v. pel 




sea. cor 



2mm 



Fig. 2. Ichthyotringa f areata (Agassiz). Pectoral girdle of the right side in lateral view. 

Paired fins and girdles. The pectoral girdle is shown in lateral view in Text- 
figure 2. The post-temporal bears a flattened narrow dorsal limb in association with 
the occipital region of the skull. The supracleithrum articulates with the postero- 
medial region of the post-temporal, and is a short but expanded bone inclined slightly 
backwards ventrally. The cleithrum is a large curved bone, whose anterior edge is 
turned medially to form the posterior border to the branchial opening. Two post- 
cleithra are present, attached to the posterior region of the supracleithrum and 
cleithrum. The dorsal postcleithrum is oval in shape, but the ventral postcleithrum 
is thinner and more elongated. The endoskeletal girdle is indistinct. The coracoid, 
however, is expanded where it curves forwards to connect with the tip of the cleithrum. 
The pectoral fin articulates with the posterior edge of the endoskeletal girdle through 
the intermediary of four hour-glass shaped radials. The pectoral fins are large and 
appear to have a variable number of fin rays, between 16 and 20. The rays are seg- 
mented and bifurcated distally. 

The pelvic girdle consists of a pair of elongated triangular bones lying in the mid- 
ventral region of the body wall. The bones contact each other at their extreme 
anterior ends and posteriorly by medial bony processes behind the fin insertion, this 
being in advance of the origin of the dorsal fin. Each fin consists of approximately 
n rays, similar in form to those of the pectoral. 



ESPECIALLY MYCTOPHOIDS 



<0 




o 
+J 

4) 



« 






I 



i- 



6 



12 UPPER CRETACEOUS TELEOSTS 

Vertebral column. The entire vertebral column is shown in the restoration of the 
skeleton (Text-fig. 3). It consists of 40 vertebrae of which approximately 23 are 
precaudal and 17 caudal. The centra are slightly longer than they are deep and 
are considerably mesially constricted. Laterally the centra bear about three 
longitudinal raised ridges, but more posteriorly in the caudal region the lateral 
ridges are reduced to two. In the precaudal region neural spines are very fine and 
long with a prominent backward curvature, but in the caudal region the spines 
become shorter, stouter and straighter. All of the neural arches are intimately 
connected to the corresponding centra. The precaudal vertebrae bear small antero- 
ventral transverse processes with which pleural ribs articulate. The ribs are long, 
thin and strongly curved both posteriorly and ventrally around the body cavity. 
The enlargement and ventral prolongation of the transverse processes begins at 
about the 20th precaudal vertebra and fusion occurs by the 24th vertebra. The 
haemal arches thus produced gradually become drawn out into backwardly project- 
ing haemal spines. Prezygapophyses and postzygapophyses are not prominent on 
the precaudal vertebrae but are more apparent caudally. 

Intermuscular bones, both epineurals and epipleurals, occur on the first 33 verte- 
brae. Anteriorly the epineurals articulate with the neural arches, but more post- 
eriorly the region of attachment rises on to the neural spines. Epipleurals anteriorly 
are associated with the centra, but in the caudal region associate with the haemal 
arches and spines. 

Median fins and tail. The fins are shown in the restoration of the whole skeleton 
(Text-fig. 3). The dorsal fin is made up of approximately 13 rays and is situated 
about mid-way between the occiput and the base of the caudal fin. The first ray is 
short and unbranched unlike the remainder which are longer and branched. All of 
the fin-rays are segmented. The proximal radials are long and thin with expanded 
head regions. The anteriormost radial has an expanded shaft region which bears a 
median anterior keel-like expansion. The distal ends of the proximal radials are 
expanded antero-posteriorly and inclined at a slight angle to the long axis of the fin ; 
this is probably due to the fusion of the medial radials to the proximal radials. 

The anal fin is composed of 15 rays, situated behind the level of the posterior end of 
the dorsal fin. None of the proximal radials are expanded, all are thin and rod-like 
with only slightly expanded heads. The first ray is short and unbranched, the re- 
mainder are longer and branched. 

The caudal skeleton (Text-fig. 4) consists of six vertebrae, four preural vertebrae 
and two ural vertebrae, the second ural merely being a terminal half-centrum. 
Preural vertebrae 2, 3 and 4 all bear prominent neural and haemal spines which 
project more noticeably backwards than the spines of the vertebrae preceding them. 
The spines of these three preural vertebrae support the small accessory rays of the 
caudal fin. The first preural centrum is reduced in length and inclined slightly 
postero-dorsally. The haemal spine of preural 1 (the parhypural) is prominent and 
expanded. Dorsally the first preural vertebra does not bear a neural spine, simply an 
enlarged neural arch. The first ural vertebra is even more reduced in extent than the 
first preural vertebra and is upturned. On its ventral edge the first ural vertebra 
bears two expanded hypurals which extend to the mid-line of the caudal fin. Ural 



ESPECIALLY MYCTOPHOIDS 



13 



vertebra 2 is a small fraction of a centrum bearing the upper four hypurals. As- 
sociated with the dorsal and lateral regions of the first preural vertebra and both ural 
vertebrae, are several uroneural elements which are long and thin and inclined 
postero-dorsally. The first uroneural, which is clearly associated with preural 
vertebra 1 and ural vertebra 1, bears a median anterior expansion. This expansion 
extends anteriorly above the neural arch of the first preural vertebra towards the 
neural spine of the second preural vertebra. With the dorsal region of this uro- 
neural expansion at least two epurals are associated. 



pu4 




Fig. 4. Ichthyotringa furcata (Agassiz). Caudal fin skeleton in lateral view. 



The caudal fin is composed of 19 principal rays of which 17 are branched. The 
accessory rays are shorter and unbranched and occur both above and below the 
caudal peduncle, there being 16 accessory rays in all. 

Scales. The whole body is covered with small cycloid scales, all of which are 
marked with concentric ridges. The scales along the course of the lateral line are 
noticeably different, being squarer in outline and bearing a longitudinal raised ridge. 



i 4 UPPER CRETACEOUS TELEOSTS 

Ichthyotringa delicata (Hay) 

(Text-fig. 5) 

1903 Rhinelhts delicatus Hay : 432. 

Diagnosis. Ichthyotringa of estimated standard length 30 mm. Head shallow 
with a flattened roof. Mandibular articulation lying below the orbit. Pre- 
operculum expanded antero-ventrally but with posterior edge remaining vertical. 
Vertebral column consisting of 43 vertebrae. 

Holotype. Specimen number 4530 in the American Museum of Natural History, 
from the Middle Cenomanian of Hajula in the Lebanon. 

Material. B.M.N.H. specimen number P. 48831. This specimen was obtained 
during the transfer development of further specimens of a different genus. P.48831 is 
a small, fragmentary, flattened specimen entirely dissociated from the matrix. The 
snout region is absent and the body is broken and incomplete but the skull-roof is 
intact up to the anterior end of the orbits. 

Remarks. Due to the fragmentary nature of the specimen a complete description 
is not available, but those features which are observable warrant its reception in the 
genus Ichthyotringa. Apart from the difference in age (P. 48831 is from Hakel) one 
feature separates it from Ichthyotringa jurcata found in the Sahel Alma deposits, this 
being the forward migration of the mandibular articulation to a point below the orbit. 
Hay (1903 : 432) erected a new species for a rhinellid from Hajula in the Lebanon on 
a small specimen mentioned above. The two specimens clearly belong to the same 
species. 

Description. Neurocranium. The neurocranium, without the rostral region, is 
shown in dorsal view in Text-figure 5. The frontals meet in the mid-line in a sinuous 
suture and form the major part of the skull-roof, which is flat. The frontals attain 
their greatest width at the hind end of the orbit above the sphenotics. Here they 
form the anterior half of the roof of the cranial cavity. Above the orbits the lateral 
expansions of the frontals are distinctly marked off from the medial regions, the 
demarcation between these two areas is the tube through which the anterior part 
of the supraorbital sensory canal ran. At first sight these lateral regions of the 
frontal appear to be elongated supraorbital bones. The frontals are unornamented 
except for the slightly raised ridges associated with the course of the sensory 
canal. The lateral and posterior edges of the frontal do not fuse with the neigh- 
bouring roofing bones, but overlap them. The tube which contained the supra- 
orbital sensory canal is indicated by pores opening on to the surface at various points. 
The main supraorbital sensory canal passed anteriorly above the orbit, forming a 
demarcation between medial and lateral regions of the frontal. Further subsidiary 
branches of the supraorbital sensory canal are present within the frontal; one 
branch extends medially and two posteriorly on to the surface of the frontal. Two 
further posterior branches passed on to the dorsal surface of the pterotic. The supra- 
orbital canal also appears to have connected with the infraorbital sensory canal with- 
in the frontal on the dorsal surface of the sphenotic. The otic branch of the infra- 



ESPECIALLY MYCTOPHOIDS 



15 



SOS c 




epo 



st.s.c 



inf. s.c 



o.inf. s.c 



op. s.c 



Fig. 5. Ichthyotringa delicata (Hay). Neurocranium in dorsal view. The anterior end 
of the snout is absent and the broken line on the right side of the figure indicates the 
course of the sensory canals. From B.M.N.H. P. 48831. 



orbital sensory canal passed forwards within the pterotic to enter the sphenotic. 
After contacting the supraorbital sensory canal in the frontal the infraorbital canal 
passed through the sphenotic laterally and down into the last infraorbital bone. 

The sphenotic projects laterally from the frontal. The greatest width of the neuro- 
cranial roof is between these extremities of the sphenotic extensions. The sphenotic 
also inclines ventrally. Dorso-laterally the upper surface of the sphenotic is ex- 
cavated for the reception of the last infraorbital bone. 

The pterotic forms the postero-lateral edge of the skull-roof and extends consider- 
ably on to the roof. Antero-medially the pterotic is covered by the backward exten- 
sion of the frontal, but postero-medially it overlaps the lateral edge of the parietal 
forming the roof of the post-temporal fossa. The pterotic contacts the sphenotic 
anteriorly and the epiotic posteriorly, the supratemporal lying dorsal to the junction 
with the epiotic. The dorsal surface of the pterotic is smooth but dorso-laterally 
several elongated pores open into the tube which housed the otic branch of the infra- 
orbital sensory canal. The most posterior pore is the largest and received the sensory 



16 UPPER CRETACEOUS TELEOSTS 

canal from the preoperculum. The tube in the pterotic opens posteriorly in a shallow 
groove above the opening of the post-temporal fossa. The posterior end of the fossa 
is roofed by the supratemporal bone which transmitted the lateral line canal into the 
rear end of the pterotic. 

The parietals are prominent oblong bones meeting in the mid-line in a sinuous 
suture. Anteriorly the parietals are overlapped by the frontals and laterally by the 
pterotics. The posterior edges of each parietal are obliquely arranged so that they 
meet in the mid-line in a shallow V-shaped indentation. Attached to the rear edge 
of both parietals are small transverse strips of bone each having a grooved dorsal 
surface. These slips of bone appear to be the remnants of the transverse limbs of the 
supratemporal. The main part of the supratemporals are small and positioned 
laterally. The supratemporal sensory canal ran across the skull-roof through both 
supratemporal bones and across the dorsal surfaces of the parietals and on to these 
reduced slips of bone. 

The supraoccipital is small, just appearing on the roof of the skull. The antero- 
dorsal extent of the supraoccipital is covered by the parietals. A small supra- 
occipital crest extends from the posterior face of the bone, but does not rise above the 
plane of the skull-roof. The supraoccipital contacts the epiotics laterally in a 
straight suture. 

The epiotics extend postero-laterally from the supraoccipital and bear slight dorsal 
thickenings for the articulation of the dorsal limbs of the post-temporals. The 
epiotic contacts the parietal anteriorly and forms the postero-medial border to the 
post-temporal fossa. The dorsal surface of the epiotic and the extreme posterior 
regions of both the parietal and the pterotic are covered dorsally by the supra- 
temporal. 

The parasphenoid is straight and considerably expanded laterally. At the an- 
terior end of the orbit the parasphenoid is an extremely wide flat plate of bone. The 
lateral ethmoid limits the orbit anteriorly, arising from the ventral surface of the 
frontal. Ventrally the lateral ethmoid is spread and its medial part appears to 
attach to the dorsal surface of the parasphenoid. More laterally the lateral ethmoid 
is free from the parasphenoid and attaches to the postero-dorsal part of the palatine. 

In the hind wall of the orbit the pleurosphenoid meets the sphenotic, prootic and 
the undersurface of the frontal. Joined to the anterior edge of the pleurosphenoid 
and pressed against the ventral surface of the frontal in the middle of the orbit, 
is another bone which appears to be the orbitosphenoid. Ventral to the pleuro- 
sphenoid a further slip of bone is visible and this may represent the upper end of a 
basisphenoid. 

In front of the orbit overlapping part of the frontal and lateral ethmoid there is a 
thin, small bone which may represent a nasal or a supraorbital. 

Remarks. The remainder of the description of Ichthyotringa delicata will be con- 
fined to those features which differ from Ichthyotringa furcata or which are exhibited 
better. 

The infraorbital bones, below the orbit, are crushed against the palate but show 
appreciably better than in Ichthyotringa furcata. The two infraorbitals below the 



ESPECIALLY M YCTOPHOIDS 17 

orbit are long and narrow with the second one showing a slight posterior deepening. 
Three expanded infraorbitals lie posterior to the orbit covering practically the whole 
of the quadrate, metapterygoid and hyomandibular. 

The suspensorium of the lower jaw is inclined forwards with the quadrate condyle 
situated below the hind end of the orbit, unlike Ichthyotringa furcata which has a 
vertical suspensorium. The hyomandibular curves anteriorly in its lower regions 
and is shallow. The quadrate is large and forms a continuation of the forward curv- 
ature of the lower end of the hyomandibular. The endopterygoid is visible and is 
connected to the medial region of the ectopterygoid. It is completely covered 
ventrally with a uniform series of small teeth. The teeth on the ectopterygoid are 
larger and appear to be slightly recurved. 

The preoperculum is considerably larger than in Ichthyotringa furcata with the 
anterior edge curving forwards ventrally following the curvature of the posterior 
edges of both hyomandibular and quadrate. The posterior edge of the preoperculum 
remains vertical in position, thus the properculum has an expanded ventral region. 
The anterior edge is thickened and rounded but more posteriorly it is thinner. 
The tube which housed the preopercular sensory canal has several openings to 
the external surface throughout its length. The suboperculum is well exhibited 
with its anterior edge thickened and extended dorsally as a prominent spine lying 
medial to the operculum. This antero-dorsal spine is as deep as the main body of 
the suboperculum itself. 

Part of the hyoid arch is visible with a slightly elongated, shallow ceratohyal, fol- 
lowed by a shorter deeper epihyal. A prominent hypohyal lies anterior to the cerato- 
hyal, and a median urohyal extends backwards from the region of conjunction of the 
hypohyals. At least 8 branchiostegal rays are present (9 appear in Ichthyotringa 
furcata) but there may have been more. The first five articulate with the ceratohyal, 
the remainder with the epihyal. The branchiostegals gradually increase in thickness 
from the anterior region but they do not increase in length. 

What can be seen of the rest of the body is in agreement with the features already 
described in Ichthyotringa furcata. 



Incertae sedis 

Genus APATEODUS Woodward, 1901 

Diagnosis (emended). Head little elongated but large and robust. Post- 
temporal fossa completely roofed. Parietals only just contact each other in the mid- 
line. Supraoccipital large. Orbitosphenoid and basisphenoid both present. 
Jugular canal present in the prootic. Posterior infraorbitals greatly expanded. 
Maxilla untoothed, forming two-thirds of the gape. Premaxilla laminate and 
toothed. Large supraorbital bone present. Mesocoracoid arch in the pectoral 
girdle. 

Type species. Pachyrhizodus glyphodus Blake. 



18 UPPER CRETACEOUS TELEOSTS 

Remarks. The genus Apateodus is only known by fragmentary material of the 
head region, thus making it difficult to place with any certainty. Its position close 
to the genus Ichthyotringa will be discussed in the consideration of the suborder 
Ichthyotringoidei as a whole (pp. 169-176). 

Apateodus striatus Woodward 
(Text-figs. 6-12) 

1837 ? Sanrocephalus striatus Agassiz, 5 : 1 : pi. 25c, figs. 17-20. 

1844 ? Saurocephalus striatus Agassiz, 5 ; 1 : 102. 

1887 Enchodus corneti Forir : 37, pi. 1, figs. 2, 3. 

1901 Enchodus corneti Forir ; Woodward : 204 (name only). 
iqot Apateodus striatus Woodward : 260, pi. 13, fig. 6, pi. 14. 

1902 Apateodus striatus Woodward ; Woodward : 38, pi. 11, figs. 1-7, text-fig. 9. 
1912 Apateodus striatus Woodward ; Woodward : 246, pi. 54, fig. 4. 



1924 Apateodus corneti (Forir) 
1929 Apateodus corneti (Forir) 
1952 Apateodus corneti (Forir) 



Kruizinga : 293. 
Leriche : 276. 
Kruizinga : 42. 



Diagnosis (emended). Apateodus in which the head is large, with a pointed 
snout. Premaxillae small, toothed and laminate. Maxillae long and rod-like. 
Maximum depth of the mandible equalling one-fifth of its length. Large teeth on 
the palatines, ectopterygoid and dentary ; teeth much laterally compressed and 
marked with fine longitudinal striations. 

Holotype. B.M.N. H. No. 49821, a small skull from the Turonian of Lewes in 
south-east England. 

Material. The holotype and specimen number 26241 in the Museum of the 
Institute of Geological Sciences, London : the latter specimen was prepared in acetic 
acid. This fine head had previously been described and figured by Woodward 
(1912 : 246, pi. 54, fig. 4). Several specimens were examined in the Musee royal 
d'Histoire naturelle, Brussels, these being from the Maastrichtian deposits in 
Belgium and Holland. 

Description. Neurocranium. The neurocranium is shown in dorsal, ventral, 
lateral and posterior views in Text-figures 6-9. Posteriorly the neurocranium is 
broader than it is deep and gradually shallows anteriorly. The front of the neuro- 
cranium is acutely pointed, and the orbit occupies the central one-third of its length. 

The f rentals form the major part of the neurocranial roof, meeting in the mid-line 
in a sinuous suture. Posteriorly the frontals contact the parietals, also in a sinuous 
suture. The frontals are flat medially, and relatively unornamented. More later- 
ally they bear several shallow longitudinal ridges which are more concentrated above 
the region of the supraorbital sensory canal. A shallow ridge passes anteriorly 
from the centre of ossification above the hind end of the orbit, and demarcates the 
lateral frontal region. This lateral region is curved to form the upper border of the 
orbit. The shallow ridge continues anteriorly to form the lateral edge of the neuro- 
cranium. The frontal above the orbit is curved, and anteriorly, on a level with the 
lateral ethmoid, there is a large supraorbital. The supraorbital runs antero-ventrally 



ESPECIALLY MYCTOPHOIDS 



19 



continuing the curvature of the frontal around the orbit. A shallow ridge passes 
postero-medially from the centre of ossification on the frontal and this is continued 
back on the surface of the parietal. Several small, short ridges extend medially 



so.s.c 



inf.sc 




Fig. 6. Apateodus striatus Woodward. Neurocranium in dorsal view. The broken line 
on the left side of the figure indicates the course of the sensory canals. From the 
collection of the Institute of Geological Sciences, number 26241. 



20 UPPER CRETACEOUS TELEOSTS 

from the centre of ossification. Postero-laterally the frontal meets the sphenotic 
above the hind end of the orbit and posterior to this is overlapped by the medial 
extension of the anterior region of the pterotic. The extreme anterior extent of the 
pterotic roofs the anterior end of the post-temporal fossa, while the frontal also 
contributes to the antero-medial walls and roof of the fossa. The ventral surface of 
the frontals bears a broad median channel which is bridged posteriorly by the orbito- 
sphenoid and the pleurosphenoids. 

The parietals are partially separated in the mid-line by the expansion of the supra- 
occipital, but meet anterior to the supraoccipital in a short suture. Each parietal 
bears a shallow, transverse crest which contacts its fellow in the mid-line. The 
lateral extremities of these crests join with the ridges passing back on to the parietals 
from the frontals. The combined crests so produced pass postero-laterally from the 
parietals on to the epiotics, terminating on the rear edges of the neurocranium above 
the post-temporal fossae. Laterally the parietal is overlapped by the medial ex- 
tension of the pterotic in the roof of the fossa. 

The supraoccipital is large and appears on both the posterior and dorsal faces of the 
skull. Laterally the supraoccipital meets the epiotics on both the dorsal and post- 
erior surfaces of the neurocranium. Ventrally on the posterior face of the skull the 
supraoccipital contacts the exoccipitals. From the rear face of the supraoccipital 
a moderately large crest projects posteriorly but does not rise above the plane of the 
skull-roof. 

The epiotics form part of the occipital border, contacting the parietals anteriorly 
and being overlapped by the medial extension of the pterotic. On the posterior face 
of the neurocranium the epiotics contact the exoccipitals ventrally. The lateral sur- 
face of the epiotic forms the posterior part of the medial wall, roof and floor of the 
post-temporal fossa. The fossa is complete and opens by a discrete fenestra on the 
posterior face of the neurocranium. Its roof and medial wall are composed of frontal, 
parietal and epiotic while the lateral part of the fossa seems to be composed solely of 
pterotic, but there may possibly be inclusions from the sphenotic and the prootic in 
the anterior region of the fossa. On the dorsal surface of the epiotic just medial to 
the post-temporal fossa, the epiotic is thickened into a slight knob which provides the 
articulatory facet for the dorsal limb of the post-temporal. 

The sphenotic contacts the frontal laterally and forms the upper posterior border 
to the orbit. The dorsal surface of the sphenotic is excavated and smooth and forms 
the anterior part of the ill-defined dilatator fossa. Near to the lateral edge of the 
sphenotic a small foramen may have transmitted a branch of the otic nerve. 

The pterotic joins the sphenotic anteriorly in the dilatator fossa and extends 
medially above the post-temporal fossa. The pterotic inclines ventro-laterally, and 
has a shallow postero-laterally directed crest along its entire length. 

The supraorbital sensory canal ran within the frontals anteriorly above the orbit 
and posteriorly back towards the parietal. Several subsidiary branches of the main 
canal passed on to the smooth medial region of the frontal and there might have 
been a complete frontal commissure, as in Aulopus (p. 204), and many other mycto- 
phiforms (Gosline, Marshall and Mead, 1966 : 7). The lateral line canal entered 



ESPECIALLY MYCTOPHOIDS 



a.t.fc 



f. i.c.a 




f.o.a 



20mm 



Fig. 7. Apateodus striatus Woodward. Neurocranium in ventral view. From the 
collection of the Institute of Geological Sciences, number 26241. 



22 UPPER CRETACEOUS TELEOSTS 

the posterior end of the pterotic through a small pore and continued anteriorly 
within the pterotic as the otic branch of the infraorbital sensory canal. The 
main infraorbital sensory canal traversed the antero-dorsal region of the sphen- 
otic and passed into the frontal, where it contacted the otic branch. The infraorbital 
canal and the supraorbital sensory canal may have contacted each other within the 
frontal. A supratemporal sensory canal crossed the roof of the skull on the parietal 
bones and entered the pterotic laterally above the post-temporal fossa. The pterotic 
in this region bears several flattened pores through one of which the supratemporal 
canal appears to have passed. Within the pterotic the supratemporal canal may 
have joined the otic branch of the infraorbital sensory canal. 

The mesethmoid appears to be small, forming the anteriormost region of the snout. 
The mesethmoid contacts the vomer, and their combined lateral surfaces provide the 
articulatory region for the internal face of the palatine. 

The lateral ethmoids are stout, well ossified bones firmly attached to the underside 
of the frontals and forming the anterior boundary of the orbit. Ventrally the lateral 
ethmoid is very irregular in outline, this area being filled with cartilage and providing 
the posterior suspensorial point for the palatine. 

The vomer is an extensive untoothed bone, attached to the mesethmoid anteriorly, 
and extending back below the parasphenoid as a broad flat plate. 

The parasphenoid is long and curved and anteriorly overlies the posterior shaft of 
the vomer. Below the orbit the dorsal surface of the parasphenoid bears a median 
longitudinal groove ending at the basisphenoid pedicel. This groove presumably 
served for the insertion of the ventral edge of a membranous interorbital septum. 
The ascending processes of the parasphenoid arise just posterior to the basisphenoid 
pedicel and pass vertically upwards to contact the prootics. The foramen through 
which the internal carotid artery passed lies at the base of the ascending process. 
Posteriorly the parasphenoid is reduced in depth and is flattened ventrally, where it 
contacts the ventro-lateral edges of the basioccipital. The parasphenoid ends below 
the occipital condyle, and the myodome opens posteriorly below the condyle. 

The basioccipital forms the lower half of the occipital condyle and contacts the 
exoccipitals above. The occipital condyle is practically square in outline with 
a prominent notochordal pit near to the dorsal edge of the basioccipital. The 
basioccipital, at its anterior end, contacts the prootic and the ascending process of the 
parasphenoid. Ventrally the basioccipital is grooved where it forms the roof of the 
posterior myodome. Just anterior to the condyle on the lateral face of the basi- 
occipital is a group of pits and ridges which may have marked the insertion of a 
ligament from the supracleithral region of the pectoral girdle but this is by no means 
certain. Internally the basioccipital forms the posterior parts of the otolith chambers 
and the posterior part of the myodome roof. The basioccipital above the myodome 
bears a prominent median longitudinal ridge which partially separates the otolith 
chambers. The basioccipital is inclined antero-dorsally above the myodome and 
contacts the posterior edges of the prootics. The otolith chambers are continued as 
recesses on the medial face of the prootics. 



ESPECIALLY MYCTOPHOIDS 



23 



The prootics are large, complex bones containing the trigemino-facialis chamber. 
Ventrally the prootic consists of a single sheet of bone contacting the ascending 
process of the parasphenoid. More dorsally it divides into two sheets, the inner one 
curving medially to meet its partner of the opposite side in the mid-line to form the 
prootic bridge, while the lateral sheet passes dorso-laterally to contact the sphenotic 
and the pterotic. The lateral sheet is produced medially above the prootic bridge to 
form the rear face of the orbit. The otolith chambers are closed anteriorly by the 




f. i.c a 



fo.a 



Fig. 8. Apateodus striatus Woodward. Neurocranium in lateral view. From the 
collection of the Institute of Geological Sciences, number 26241. 

junction of the prootic bridge with the lateral wall of the prootic. The basioccipital 
is joined to the hind edge of the prootic bridge and thus separates the cranial cavity 
from the myodome. At about the middle of its length, the prootic bridge is pierced 
near its lateral edge, just internal to the anterior extension of the otolith chamber, by 
a foramen which transmitted the abducens nerve into the myodome. Above the 
prootic bridge the anterior, orbital face of the prootic is irregular in shape and deeply 
pitted. This part of the prootic meets the sphenotic dorso-laterally, and the pleuro- 
sphenoid and basisphenoid medially. The lateral face of the prootic contacts the 
sphenotic and the pterotic within the hyomandibular facet. 

The trigemino-facialis chamber is divided by a bony septum into a medial pars 
ganglionaris and a lateral pars jugularis. The pars jugularis is a long horizontal 
canal lying within the prootic, lateral to the pars ganglionaris. The canal opens 
posteriorly near to the rear edge of the prootic through a small elongated foramen. 
The canal widens anteriorly and opens into the rear face of the orbit. There are two 
foramina leading from the pars ganglionaris into the pars jugularis. The more 
posterior foramen is the facial foramen which transmitted the hyomandibular 
branch of the facial nerve. The more anterior foramen is the trigeminal and trans- 
mitted the trigeminal, buccal, otic, superficial ophthalmic and profundus nerves. 
The palatine nerve did not enter the pars jugularis but instead passed ventrally 



24 



UPPER CRETACEOUS TELEOSTS 



through a small foramen medial to the facial foramen into a canal which opens 
ventrally into the myodome. The posterior opening of the jugular canal only trans- 
mitted the jugular vein. Dorso-lateral to the pars jugularis a large foramen trans- 
mitted the hyomandibular nerve. Ventro-laterally a small foramen leads upwards 
through a canal into the pars jugularis, and this carried the orbital artery. Through 
the anterior opening of the pars jugularis the jugular vein, the orbital artery, and the 
buccal branches of both the trigeminal and the facial nerves were transmitted. Two 
very small foramina are also found in the lateral wall of the pars jugularis and may 
have transmitted a small artery and vein into the muscle masses next to the prootic. 
Several additional foramina are found in the hind wall of the orbit. The otic 
nerves from the anterior opening of the pars jugularis passed upwards into the 
sphenotic through several elongated foramina. Medial to the anterior opening of the 
pars jugularis a large foramen transmitted the superficial ophthalmic nerves, the 
profundus, and probably a mandibular branch of the facial or the trigeminal. The 
superficial ophthalmic nerves passed dorsally on to the face of the pleurosphenoid 
where their course is marked by a groove. A further small foramen in the hind wall 
of the orbit dorso-medial to the other foramina may also have transmitted super- 
ficial ophthalmic nerves. The oculomotor passed through a foramen lying ventro- 
lateral to the suture between the basisphenoid and the prootic. 



soc 



fd.pt 




par 



20mm 



Fig. a. 



Apateodus striatus Woodward. Neurocranium in posterior view, 
collection of the Institute of Geological Sciences, number 26241. 



From the 



The exoccipitals meet in the mid-line on the posterior face of the neurocranium and 
enclose the foramen magnum. Each exoccipital forms part of the occipital condyle 
together with the basioccipital. The exoccipitals contact the supraoccipital dorsally, 
the epiotics laterally and the intercalar ventro-laterally. On either side of the fora- 
men magnum are two small foramina which transmitted occipital nerves to the 
vertebral column. On the lateral face of the skull the exoccipital meets the basi- 
occipital ventrally, the prootic anteriorly, the pterotic dorsally and the intercalar 



ESPECIALLY MYCTOPHOIDS 25 

postero-dorsally. A large foramen in the lateral wall of the exoccipital transmitted 
the vagus nerve. Anterior to the vagus foramen is a further smaller foramen for the 
glossopharyngeal nerve. 

The pterotic makes up the posterior two-thirds of the hyomandibular facet and 
contacts the prootic and sphenotic anteriorly within the hyomandibular facet. 
Below the facet, the pterotic contacts the prootic, exoccipital and intercalar. 

The intercalar is small and superficial, lying ventral to the posterior opening of the 
post-temporal fossa. The ventral surface of the intercalar is flattened, but dorsally 
provides a concave facet for the articulation of the ventral limb of the post-temporal. 

The sphenotic is large and forms the postero-dorsal corner of the orbit. Within 
the orbit the sphenotic meets the prootic ventrally and the pleurosphenoid medially. 

The basisphenoid is Y-shaped with a long pedicel which presumably contacted the 
parasphenoid ventrally, although not evident in the specimen. The upper arms of 
the Y are spread out and articulate with the medial edges of the prootic above the 
prootic bridge. These dorsal arms separate the hypophysial fenestra ventrally from 
the optic fenestra dorsally. The basisphenoid also articulates dorsally with the medial 
edges of the pleurosphenoids. The basisphenoid pedicel bisects the opening of the 
posterior myodome, and gave insertion to the posterior edge of the presumably 
membranous interorbital septum. 

The pleurosphenoids are separated in the mid-line by the optic fenestra, which is 
closed by the orbitosphenoid antero-dorsally and the basisphenoid postero-ventrally. 
The pleurosphenoids are attached dorsally to the undersurface of the frontals and in 
this region are only superficially ossified. Along the ventral surface of the frontals 
the course of the superficial ophthalmic nerves is indicated by several small foramina 
through which branches of the nerve passed to the organs of the supraorbital sensory 
canal. 

The orbitosphenoid is a large, median, well ossified bone lying anterior to the 
pleurosphenoids. It contacts the undersurfaces of the frontals dorsally and the 
olfactory nerves passed through a median foramen in its anterior edge. 

Infraorbital bones. The infraorbital bones are shown in lateral view in Text- 
figure 11. A low elongated lachrymal lies at the anterior end of the orbit, followed by 
a second shallow infraorbital. The infraorbitals posterior to the orbit are greatly 
expanded (see also Woodward, 1912, pi. 54, fig. 4) extending back lateral to the 
preoperculum. 

Hyopalatine bones. The hyopalatine bones are shown in medial view in Text- 
figure 10. The hyomandibular is short, broad and forwardly inclined ventrally, 
with a single elongated head. The opercular process is prominent and projects 
posteriorly from the upper part of the rear edge of the hyomandibular. On the lateral 
face a large crest, originating near the posterior region of the head, passes ventrally 
along the posterior edge of the bone, where it ends abruptly. The upper anterior 
edge of the preoperculum rests against the rear edge of this crest. Ventrally the 
hyomandibular tapers to a stout rod of bone. Anteriorly the hyomandibular is thin 
and covered laterally by the postero-dorsal part of the metapterygoid. On the 
medial face there is a large antero-dorsal foramen through which the hyoman- 



2f> 



UPPER CRETACEOUS TELEOSTS 



dibular nerve passed. Within the bone the nerve divided ; the opercular nerve 
left through a small foramen lateral to the opercular process ; the hyoidean and 
mandibular nerves left through a large foramen in the angle between the base of the 
opercular process and the lateral crest. There is a possibility that a second branch of 
the mandibular nerve passed through a foramen on the lateral face of the hyo- 
mandibular near to the ventral extremity. 



hmVll 



opphm 




g.sym 



Fig. io. Apateodus striatus Woodward. Hyopalatine and opercular bones and mandible 
of the right side in medial view. From the collection of the Institute of Geological 
Sciences, number 26241. 



The quadrate is large, triangular and has a prominent transversely arranged, mesi- 
ally constricted condyle. The groove on the medial face of the quadrate which 
housed the symplectic is long and shallow and terminates just above the condyle. 
Posterior to the symplectic groove the postero-dorsal part of the quadrate is emarg- 
inated. Anteriorly the quadrate is flattened and extends forwards over the lateral 
face of the posterior end of the ectopterygoid. On the medial face the ectopterygoid 
extends back almost as far as the condyle. 

The metapterygoid is a thin membranous bone overlying the edges of both the 
hyomandibular and the endopterygoid but not reaching the dorsal region of the 
quadrate. Laterally the metapterygoid bears a slight antero-posteriorly orientated 
crest which forms the leading edge of the metapterygoid ventrally. 

The ectopterygoid is laterally compressed posteriorly, but widens anteriorly where 
it supports teeth. Anteriorly the ectopterygoid joins the palatine, and the medial 
edge of the ectopterygoid supports the endopterygoid. There are seven large teeth 



ESPECIALLY MYCTOPHOIDS 



27 



which decrease in size posteriorly. Each tooth is laterally compressed and acutely 
pointed with longitudinal striations more noticeable on the unexpanded basal 
regions (see also Woodward 1912 : 41, fig. 9, and Kruizinga 1924). 

The palatine is prominent and overlaps the anterior end of the ectopterygoid both 
laterally and medially. Both medial and lateral faces of the palatine are flattened 
and the ventral face of the bone is concave. The ventral concavity supports two 
prominent teeth, the large one arising at the extreme posterior end of the palatine. 
The anterior tooth is smaller and much nearer the anterior extremity. The teeth are 
identical in shape and form to those on the ectopterygoid with which they form a 
continuous row. The postero-dorsal edge of the palatine is excavated for the re- 
ception of the lateral ethmoid. The medial edge of the palatine articulates with the 
lateral edges of both the mesethmoid and the vomer. 

The endopterygoid is thin and inclined dorso-medially. Its ventral surface is 
convex and bears a large, oval patch of minute teeth. 

Dermal upper jaw. The dermal upper jaw is shown in lateral view in the recon- 
structed skull, Text-figure n. This is very incompletely preserved. The pre- 
maxilla is a small, thin lamina of bone lying alongside the anterior end of the palatine. 



hm.h 



pal.t 




dent 



ang 



Fig. 11. Apateodus striatus Woodward. Restoration of the skull in lateral view. 



The upper edge of the premaxilla is rounded but its toothed oral margin is straight. 
The teeth are minute and the premaxilla extends from the tip of the snout to just 
behind the level of the lateral ethmoid, thus occupying approximately one-third of 
the total gape. The maxilla is long, thin and narrow and does not appear to have 



28 UPPER CRETACEOUS TELEOSTS 

borne teeth. Anteriorly the maxilla lies medial to the premaxilla where it is as- 
sociated with the palatine. Posteriorly the maxilla enters the gape behind the pre- 
maxilla. In his original description Woodward (1901 : 260) identified the palatine 
as the premaxilla and the ectopterygoid as the maxilla. Later Woodward (1912 : 
246), using the specimen considered here, identified the bones correctly. 

Mandible. The mandible is shown in medial view in Text-figure io, and lateral 
view in Text-figure 11. It is long and not particularly deepened. The dentary 
forms the whole of the oral border of the bone and almost all the ventral border. 
The posterior edge of the dentary is deeply indented and has a tube opening at the 
posterior end which is continued as a groove on the face of the articular. This tube 
and groove housed the mandibular sensory canal. The oral edge of the dentary 
is thicker and stouter than the remainder of the bone and bears teeth. The teeth 
are in a single row, of unequal size, and irregular distribution. There are approxi- 
mately 7 acutely pointed, recurved teeth. The teeth are laterally compressed 
and arranged at a slightly oblique angle to the long axis of the mandible. The three 
anterior teeth are the largest and occur a short way back from the symphysial 
region. These three large teeth appear to occupy the gap between the two palatine 
teeth when the jaws are shut. Behind these large teeth, three or four small teeth 
occur and posterior to these, one or two somewhat larger teeth. The outer face of 
the dentary is unornamented although several longitudinal ridges are present above 
the mandibular sensory canal. 

The articular facet for the reception of the condyle is concave and transversely 
orientated. The facet is divided mesially by a ridge of bone corresponding to the 
mesial constriction of the condyle. The articular has a slight retroarticular process 
which is laterally grooved for the reception of the mandibular sensory canal. 

The angular is a small slip of bone applied to the postero-ventral surface of the 
articular below the retroarticular process. 

Opercular bones. The opercular bones are shown in medial and lateral views in 
Text-figures 10 and 11. The operculum is large and deep with a concave anterior 
edge. Posteriorly it is rounded but the ventral edge is obliquely inclined. The 
outer face of the operculum is smooth and unornamented. The opercular facet is 
large and in the uppermost region. The facet extends medially and is supported by 
a strengthening ridge which passes back horizontally across the internal face of the 
operculum. 

The suboperculum is prominent and lies internal to the ventral edge of the oper- 
culum. The posterior and ventral edges of the suboperculum are thin and 
rounded. 

The interoperculum is roughly oval in shape with thickened anterior and ventral 
edges. The interoperculum lies medial to the ventral preopercular region. 

The preoperculum is narrow and inclines forwards ventrally. Its anterior edge 
rests against the posterior face of the hyomandibular crest and the posterior edge of 
the quadrate. The anterior region of the preoperculum is thickened and contains a 
deep groove overhung by a flange for the preopercular sensory canal. 



ESPECIALLY M YCTOPHOIDS 



29 



Pectoral girdle and fin. The incomplete pectoral girdle is shown in medial view in 
Text-figure 12. The cleithrum is a large sigmoid bone with a flattened lateral extent 
and a medially inclined anterior edge. 

The scapula is irregularly shaped with a large scapular foramen near its centre. 
On the posterior edge of the scapula is a large saddle-shaped facet for the articula- 
tion of the anterior fin ray. Ventral to this facet the scapula edge is excavated 
longitudinally forming a facet for the reception of the fin radials. 

The coracoid is larger than the scapula and antero-laterally is attached to a 
medial flange on the internal face of the cleithrum. The coracoid is produced antero- 
ventrally into a stout flattened process. 

A mesocoracoid arch connects the coracoid with the scapula and cleithrum 
dorsally. The scapula is hollowed out medially to form a well marked canal between 
it and the mesocoracoid through which passed the dorso-medial muscles of the fin. 
The mesocoracoid is narrow, with a deep excavation mid-way along its posterior edge. 

The pectoral fin consists of at least 15 large, stout rays. Only the proximal parts 
of the rays are preserved and they are slightly flattened. 




sea. for 



Fig. 12. Apateodus striatus Woodward. Part of the pectoral girdle of the left side in 
medial view. From the collection of the Institute of Geological Sciences, number 
26241. 



30 UPPER CRETACEOUS TELEOSTS 

Family APATEOPHOLIDAE nov. 

Diagnosis. Head elongated and deepened posteriorly ; body deepened but not 
elongated. Post-temporal fossa unroofed. Parietals separated by the supra- 
occipital. Orbitosphenoid and basisphenoid absent. Posterior infraorbitals un- 
expanded. Mandibular suspensorium inclined forwards ventrally. Preoperculum 
with prominent postero-ventral spine. Vertebral elements incompletely fused ; 
centra as long as deep. Pelvics abdominal, below dorsal fin. Hyomandibular with 
a double head. Body naked except for scales along the lateral line. Dermal bones 
ornamented with minute bony tubercles. 

Genus APATEOPHOLIS Woodward, 1891 

Diagnosis (emended). As for the family, only genus. 

Type species. Rhinellus laniatus Davis. 

Remarks. This species was originally placed in the genus Rhinellus by Davis 
(1887 : 612). Woodward (1888b : 355) later placed the species in the genus Belono- 
stomus and later still in the new genus Apateopholis (1891 : 634). Ten years later 
Woodward (1901 : 232) considered Apateopholis to be synonymous with Prionolepis. 
In the present work the genus Apateopholis is re-erected to contain a single species. 

Apateopholis laniatus (Davis) 

(Text-figs. 13-15) 

1887 Rhinellus laniatus Davis : 612, pi. 37, fig. 1. 

1887 Rhinellus longirostris Davis : 611, pi. 37, fig. 3. 

1888b Belonostomus laniatus (Davis) Woodward : 355. 

1 891 Apateopholis laniatus (Davis) Woodward : 634, pi. 55, fig. 11. 

1901 Prionolepis laniatus (Davis) Woodward : 232. 

Diagnosis (emended). Apateopholis reaching 10 cm. in standard length. Head 
with opercular apparatus occupies more than one-third of the standard length. 
Dorsal fin with 14 rays situated midway along the back. Anal fin with approxim- 
ately 11 rays, remote in position. Pelvics with 10 rays, opposite the middle of 
the dorsal fin. Pectoral fins low on the flanks with 16 rays. Lateral line scales not 
overlapping. 

Holotype. B.M.N.H. specimen number P. 4745, from the Middle Cenomanian, 
Hakel, Lebanon. 

Material. The holotype and specimens in the B.M.N.H., numbers P.4869, 
P. 4026, P. 4870, the latter two having been prepared in acetic acid by the transfer 
method. 

Remarks. The species is only known by these few specimens, but it is difficult to 
understand why Woodward (1888b, 1901) placed the species in either of the genera 
Belonostomus or Prionolepis, since both of these genera possess enlarged shield-like 
flank scutes. None of the specimens listed above shows any evidence of such en- 
larged scutes, in fact they are devoid of scales, except along the lateral line. It has 



ESPECIALLY MYCTOPHOIDS 



3i 



been found convenient to re-erect the genus Apateopholis and to ally it with the 
ichthyotringids, but the differences are such as to warrant its removal into a 
separate family. 

Description. Neurocranium. The neurocranium is shown in dorsal view in 
Text-figure 13. In general shape the neurocranium is long, narrow and deepened 
posteriorly with a shortened cranial cavity. The anterior region of the neurocranium 
is drawn out into an extremely long projection which forms the basis of the rostral 
region. The length of the neurocranium is approximately half that of the body from 
the occiput to the caudal peduncle. The orbits are enormous, their diameter 



10mm 




Fig. 13. Apateopholis laniatus (Davis). Neurocranium in dorsal view. 



being much in excess of the postorbital length of the head. The frontals form the 
majority of the skull-roof and above the orbit the lateral edge bears an ornamenta- 
tion of bony tubercles. The sphenotic forms the upper posterior border to the orbit 
and extends ventro-laterally from beneath the frontal. The sphenotic joins the 
prootic ventro-laterally to form the anterior part of the hyomandibular facet. The 
lower edge of the sphenotic and the upper edge of the prootic are both expanded and 
excavated to produce two cup-shaped depressions into which the anterior region of 
the head of the hyomandibular rests. Dorsally the sphenotic contacts the pterotic in 



32 UPPER CRETACEOUS TELEOSTS 

a deep depression above the hyomandibular facet; this depression is the dilatator 
fossa. 

The pterotic forms the postero-lateral border of the skull-roof, and bears a high 
crest which arises antero-medially and terminates at the posterior end of the lateral 
edge of the bone. This crest forms the boundary between the dilatator fossa laterally 
and the post-temporal fossa medially. The post-temporal fossa is unroofed. The 
otic branch of the infraorbital sensory canal passed within the pterotic crest. Above 
the hyomandibular facet, the postero-lateral part of the pterotic crest bears two 
large pores. The larger of the two is more ventrally placed and through it passed 
the preopercular sensory canal ; the upper, smaller pore received the lateral-line 
canal from the post-temporal bone. 

The epiotic forms the posterior edge of the lateral region of the skull roof medial to 
the post-temporal fossa. On the upper region of the epiotic there is a thickened 
region with which the dorsal limb of the post-temporal articulated. The parietal 
lies in front of the epiotic and contacts the frontal anteriorly. The post-temporal 
fossa is composed medially of epiotic, parietal and frontal, and laterally of pterotic. 
These bones meet in the mid-ventral line of the post-temporal fossa. 

The parasphenoid can be seen in the base of the orbit and is expanded laterally into 
flanges which associate with the endopterygoids. Due to the deepening of the neuro- 
cranium posteriorly, the parasphenoid ascends sharply beneath the orbits to pass 
between the lateral ethmoids. 

Jaws and suspensorium. The hyomandibular is long and narrow with a well 
marked head. Anteriorly the head is expanded and rounded to fit into the cup- 
shaped depression between the sphenotic and prootic. The posterior part of the hyo- 
mandibular head is elongated, narrow and fits into the groove on the pterotic. The 
axis of the hyomandibular is inclined slightly forwards ventrally, as is the quadrate, 
thus the mandibular articulation lies below the middle of the orbit. The lateral face 
of the hyomandibular bears a large crest in its upper region. This crest arises from 
the anterior region of the head and passes sharply back to the posterior edge of the 
bone which it follows ventrally. The upper edge of the preoperculum rests against 
the rear face of this crest. Ventrally the hyomandibular tapers to end some distance 
above the quadrate. The symplectic is a long ventrally tapering bone contacting 
the hyomandibular dorsally and the quadrate ventrally. 

The metapterygoid overlaps a large part of the lateral face of the hyomandibular, 
and meets the dorsal edge of the quadrate ventrally. The metapterygoid bears an 
oblique crest near its anterior border. 

The quadrate is irregularly triangular with its posterior edge practically horizontal. 
The condyle is small and transversely orientated. The posterior edge of the quad- 
rate rests upon the anterior projection of the preoperculum. Both the metaptery- 
goid and the quadrate meet the ectopterygoid anteriorly. 

The ectopterygoid runs upwards from the quadrate at an acute angle to contact the 
palatine below the lateral ethmoids and to support the endopterygoid medially. 
The palatine is enormously elongated, extending the whole length of the snout 
region. It is relatively deep posteriorly but shallow anteriorly. The palatine is 



ESPECIALLY MYCTOPHOIDS 



33 



supported medially by the mesethmoid and vomer. Both the palatine and the ecto- 
pterygoid are furnished with many small conical teeth arranged in a single continuous 
row. 

The exact relationship of the maxilla and premaxilla to one another is uncertain. 
The premaxilla appears to form practically all of the oral edge of the upper jaw, the 
maxilla just entering the gape posteriorly. The oral border of both appears to sup- 
port a continuous row of minute teeth. 

The mandible is long, shallow and acutely pointed anteriorly. The oral edge of 
the dentary bears many small pointed teeth. The articular facet is a shallow semi- 
circular cavity with no trace of a retroarticular process behind it. Along the ventro- 
lateral surface of the dentary a deep groove is found which contained the mandibular 
sensory canal. 

The preoperculum is deep and narrow but characteristically expanded ventrally. 
The anterior edge follows the curved outline of the posterior edge of both hyomandib- 
ular and quadrate. The preoperculum ends dorsally just posterior to the upper 
limit of the crest on the lateral face of the hyomandibular. The near vertical post- 
erior edge of the preoperculum bears a row of prominent tubercles which impart a 
'toothed' appearance to the edge. Ventrally the bone is produced into a large 




v. pel 



Fig. 14. Apateopholis laniatus (Davis). Pectoral girdle of the left side in lateral view. 



34 UPPER CRETACEOUS TELEOSTS 

posteriorly projecting spine. The whole of this expanded ventral part is thickened 
and ornamented with longitudinal tuberculated ridges. The operculum is deeper than 
it is broad with the opercular facet contained in the upper half of the anterior edge. 
The suboperculum is overlapped dorsally by the operculum and is a large bone bear- 
ing several strengthening ridges on the lateral face of the anterior region. No trace 
of an interoperculum was found in the specimens examined. 

Pectoral girdle and fin. The pectoral girdle is shown in lateral view in Text-figure 
14. The post-temporal has a flattened, slightly arched dorsal limb which is orna- 
mented with longitudinal bony ridges. The long, narrow supracleithrum articulates 
with the postero-medial region of the post-temporal. The lateral line canal passed 
through the head of the supracleithrum into the post-temporal. Ventrally the 
supracleithrum lies against the lateral surface of the uppermost region of the clei- 
thrum. 

The cleithrum is large and curved, its anterior edge being inclined medially. The 
lateral face of the cleithrum is unexpanded and smooth, although bony ridges are 
present more dorsally. The endoskeletal girdle is attached to the internal face of the 
cleithrum approximately two-thirds of the way down the bone. The pectoral 
fin is thus low on the flank and consists of 16 rays, all of which are branched and 
segmented. 

Two postcleithra are also present, both being elongated and narrow. The dorsal 
postcleithrum arises as a continuation of the supracleithrum posterior to the clei- 
thrum, and the ventral postcleithrum forms a direct continuation of the dorsal one. 
Both are curved and lie internal to the fin rays. 

Median fins and tail. The median fins are shown in the restoration of the complete 
skeleton in Text-figure 15. The dorsal fin consists of approximately 14 rays and is 
situated mid-way along the back between the occiput and the base of the caudal fin. 
The first ray is unbranched, unlike the remainder which are all branched. The 
proximal radials are elongated and the first few are expanded, particularly the 
anteriormost which bears an enlarged median keel. 

The anal fin is composed of 11 rays and situated behind the level of the rear end of 
the dorsal fin. The first ray is unbranched as in the dorsal fin, but unlike the dorsal 
fin none of the proximal radials is expanded. 

The skeleton of the caudal fin is impossible to describe with any accuracy. There 
appear to be either three or four preural vertebrae and two ural vertebrae incorpor- 
ated into the fin skeleton. The number of principal caudal fin rays is 19 of which 17 
are branched. Shorter and unbranched accessory fin rays are present both above 
and below the caudal fin, approximately 20 in all. 

Vertebral column. The entire vertebral column is shown in the restoration of the 
skeleton in Text-figure 15. It is made up of approximately 43 vertebrae, of which 26 
are caudal. The individual centra are about as long as deep and slightly mesially 
constricted. The neural arches are separate ossifications to the centra, those of the 
anterior precaudal region being displaced from the centra. More posteriorly the 
neural arches are more intimately united to the corresponding centra. Prominent 



ESPECIALLY MYCTOPHOIDS 



35 



T3 



w 












> 

P 









36 UPPER CRETACEOUS TELEOSTS 

prezygapophyses and postzygapophyses are present on all of the neural arches. 
Above the anterior region of the vertebral column a noticeably expanded interneural 
element is present in the dorsal mid-line and appears to be associated with the neural 
spines of the first two precaudal vertebrae. Prominent antero-ventral transverse 
processes occur on the precaudal vertebrae with which the pleural ribs articulate. 
The ribs are long, slender, flattened and tapering. Each rib curves posteriorly and 
ventrally around the body cavity. 

The transverse processes begin to lengthen and extend ventrally at about the 14th 
precaudal vertebra. The ribs remain in articulation with the haemal arches of the 
first three or four caudal vertebrae, but these ribs are reduced in length. The haemal 
spines gradually lengthen on the first few caudal vertebrae. Ventral zygapophyses 
are present and increase in size in the caudal region. 

Intermuscular bones, both epineurals and epipleurals, are associated with the first 
25 vertebrae. The epineurals articulate with the neural arches anteriorly, but more 
posteriorly they shorten and the point of articulation is on the neural spine. An- 
teriorly the epipleurals attach to the ribs but in the caudal region they are associated 
with the haemal arches. 

Scales. The body is devoid of scales except for a single row along the lateral line. 
The scales do not overlap, and each is roughly rectangular in shape. The scales 
bear small posteriorly directed spines on their outer surfaces. On the caudal peduncle 
at the base of the caudal fin the spines become enormously enlarged and contiguous 
with one another forming a prominent flange. 

Suborder CIMOLICHTHYOIDEI 

Diagnosis. Body never deepened but elongated to a greater or lesser extent. 
Parietals separated by the supraoccipital. Post-temporal fossa either roofed or un- 
roofed, but dilatator fossa always roofed. Orbitosphenoid and basisphenoid absent. 
Supraorbital occasionally present but no antorbital. Premaxilla prominent, 
toothed or untoothed but with no ascending process. Maxilla stout, either toothed 
or untoothed but always entering gape. Teeth present on the vomer and the 
endopterygoids. Interoperculum absent ; branchiostegal rays always numerous. 
Pelvic fins always abdominal, below dorsal fin. Dorsal fin more or less extended ; 
anal fin always small and remote, near to caudal fin. Vertebral elements incom- 
pletely fused, with epineurals and epipleurals in large numbers. Vertebrae variable 
in number with always less than half being caudal. Fins without spines. Caudal 
with 19 principal rays. Squamation variable, usually restricted in extent with the 
predominant scales triradiate. 

Family CIMOLICHTHYIDAE nov. 

Diagnosis. Body fusiform, relatively shallow and only slightly elongated. 
Neurocranium shallow, acutely pointed anteriorly and only slightly elongated. 
Post-temporal fossa incompletely roofed. Jugular canal present in the prootic. 
Occipital condyle formed only of basioccipital. Maxilla stout and untoothed. 



ESPECIALLY MYCTOPHOIDS 37 

Palatine and ectopterygoid teeth barbed. Small supraorbital bone present and small 
supramaxilla. 50-60 vertebrae, approximately half being caudal. Ribs borne on 
small transverse processes. Body bearing two rows of arrow-shaped scutes on 
each flank. 

Genus CIMOLICHTHYS Leidy, 1857 

Diagnosis (emended). Cimolichthyidae in which the head is somewhat elongated 
with an acutely pointed anterior region. Palatine with two longitudinal rows of 
teeth, only innermost row barbed. Ectopterygoid teeth also barbed. Dentary with 
3 distinct rows of teeth. Dorsal fin short, mid-way along the back, consisting of 16 
rays. Anal fin shorter with 11 rays and remote in position. 

Type species. Cimolichthys levesiensis Leidy. 

Remarks. The genus was originally erected by Leidy (1857) when he proposed 
the type species Cimolichthys levesiensis, based on palatine and ectopterygoid material 
showing the characteristic barbed teeth. 



Cimolichthys levesiensis Leidy 
(Text-figs. 16-21) 

1835 Saurodon leanus Agassiz ; Feuill.: 55 {errore). 

1837 Saurodon leanus Agassiz ; 5 ; 1 : pi. 25c, figs. 30, 31 {errore). 

1844 Saurodon leanus Agassiz ; 5 ; 1 : 102 {errore). 

1850 Saurodon leanus Agassiz ; Dixon : 373, pi. 30, figs. 28, 29 ; pi. 32, fig. 10 {errore). 

1857 Cimolichthys levesiensis Leidy : 95. 

1864 Plinthophorus robustus Giinther : 115, pi. 6. 

1878 Cimolichthys levesiensis Leidy ; Newton : 790. 

1878 Cimolichthys marginatus (Reuss) Newton : 792. 

1878 Cimolichthys lewesiensis Leidy ; Dixon : 400, pi. 30, figs. 28, 29 ; pi. 32, fig. 10. 

1888a Cimolichthys levesiensis Leidy ; Woodward : 316. 

1 90 1 Cimolichthys lewesiensis Leidy ; Woodward : 221. 

1 901 Plinthophorus robustus Giinther ; Woodward : 229. 

1902 Cimolichthys lewesiensis Leidy ; Woodward : 44, pi. 9, figs. 10-14 > P^ I2 > text-fig. 10. 
1902 Plinthophorus robustus Giinther ; Woodward : 48, pi. 12, fig. 5. 

1902 Cimolichthys marginatus (Reuss) ; Leriche : 140, pi. 4, figs. 6-12. 
1906 Cimolichthys marginatus (Reuss) ; Leriche : 96, text-fig. 16. 
1949 Cimolichthys marginatus (Reuss) ; Dartevelle & Casier : 222. 

Diagnosis (emended). Cimolichthys in which the maximum width of the neuro- 
cranium slightly exceeds one-third of the length. Premaxilla ornamented with 
radiating tuberculated ridges. Length of mandible five times its greatest depth. 
Four ectopterygoid teeth, the foremost two being apically barbed. Approximately 
60 vertebrae of which 30 are caudal. Centra slightly longer than deep and mesially 
constricted. Pectoral fin with 12 rays, pelvic fin with 10 or 11 rays arising below the 
posterior region of the dorsal fin. Dorsal fin with 15 or more rays, anal fin with 10 
rays. Body naked except for two rows of imbricate, arrowhead-shaped scutes on 
each flank. 



38 UPPER CRETACEOUS TELEOSTS 

Holotype. B.M.N.H. specimen number 4039, an imperfect left ectopterygoid 
from the Chalk of Sussex. 

Material. Specimens in the B.M.N.H., notably numbers P.i8ioa and P.1811. 
The latter specimen was figured by Woodward (1902 : 45, pi. 12, fig. 1), and both 
specimens were prepared in acetic acid. The holotype of Giinther's (1864) Plintho- 
phorus robustus, number 38113 in the B.M.N.H. (figured by Woodward, 1902, pi. 12), 
was also examined and is placed in this species as Woodward (1901 : 229) tentatively 
proposed. 

Remarks. Leidy's naming of Cimolichthys levesiensis on palatine and ecto- 
pterygoid material was the first recognizable description of the species after Agassiz's 
errors in ascribing these to the genus Saurodon. Between these two references 
Reuss (1845) had described isolated teeth of a cimolichthyid as the fin spines of a 
shark. Leriche (1902) synonymized the two species under the name Cimolichthys 
marginatus (Reuss). However Leriche was definitely using Cimolichthys levesiensis 
material, whereas the earlier material used by Reuss was not so definitely attribut- 
able to this species. 

The following are the references to Cimolichthys marginatus 

1845 Spinax marginatus Reuss : 8, pi. 4, figs. 10, 11. 

1848 Acanthias marginatus (Reuss) Giebel : 301. 

1855 Anenchelum marginatum (Reuss) Hebert : 350, pi. 27, fig. 4. 

1874 Anenchelum marginatum (Reuss) ; Barrois : 131. 

1874 Trichiurides (Anenchelum) marginatum (Reuss) Winkler : 39. 

1875 Saurocephalus marginatus (Reuss) Geinitz : 226, pi. 43, fig. 38. 
1887 Saurodon marginatus (Reuss) Dames : 77. 

1897 Saurocephalus marginatus (Reuss) ; Leonhard : 69, pi. 6, fig. 17. 

Description. Neurocranium. The neurocranium is shown in dorsal, ventral, 
lateral and posterior views in Text-figures 16 to 19. It is long and shallow with the 
maximum width occurring behind the orbits. The snout is very acutely pointed 
but the neurocranium gradually becomes wider more posteriorly. 

The frontals form practically the whole of the skull-roof and are long tapering 
bones ending anteriorly by insertion on to the rear edges of the mesethmoid. Each 
frontal meets its fellow in the mid-line in a straight suture and overlaps the parietal 
and supraoccipital posteriorly. Postero-laterally the frontal is excavated above the 
sphenotic and the pterotic and only enters slightly into the roof of the post-temporal 
fossa. The surface ornamentation of the frontal is confined to the more lateral 
regions and is in the form of ridges of bone radiating from the centre of ossification 
above the sphenotic. The ridges which extend anteriorly terminate on the lateral 
edge of the frontal anterior to the orbit, those ridges which extend antero-laterally 
terminate on the edge of the frontal where it forms the upper border of the orbit. 
The ridges which extend posteriorly are arranged in a fan, each ridge ending on the 
edge of the frontal where it overlaps the other roofing bones posteriorly. 

The supraoccipital extends on to the skull-roof and forms part of the hind wall of 
the neurocranium. Anteriorly the supraoccipital is overlapped by the frontals and 



ESPECIALLY MYCTOPHOIDS 



39 



separates the parietals and the epiotics. The rear face of the supraoccipital gives 
rise to a small backwardly projecting crest which does not extend above the plane of 
the skull-roof. 




infs.c 



o.inf s c 



epo 



Fig. i 6. Cimolichthys levesiensis Leidy. Neurocranium in dorsal view. The broken line 
on the right side of the figure indicates the course of the sensory canals. From 
B.M.N.H. number P.1811. 



4 o UPPER CRETACEOUS TELEOSTS 

The parietals are narrow, transversely orientated strips of bone, which laterally 
form part of the medial wall of the post-temporal fossa, meeting the pterotic anteri- 
orly and the epiotic posteriorly. The post-temporal fossa itself is small and appears 
to be only partially roofed. Posteriorly where the fossa is unroofed it is bounded 
medially by the epiotic and the parietal, and laterally by the pterotic. 

The epiotics form part of the occipital border of the skull-roof, bounding the pari- 
etals posteriorly. On the posterior face of the neurocranium they are separated 
medially by the supraoccipital, and ventrally contact the exoccipital. Within the 
posterior region of the post-temporal fossa, the epiotic meets the pterotic in the mid- 
line of the floor of the fossa. The whole occipital border of the roof is angled sharply 
backwards from the midpoint on the supraoccipital, the epiotic forming the termina- 
tion of this backward extent. 

The sphenotic extends laterally from beneath the frontal in the posterior orbital 
region and is smooth and unornamented. A slight crest originates on the sphenotic 
and extends posteriorly to be continued on the pterotic, the whole extent of this crest 
forming the postero-lateral border of the skull-roof. 

The pterotic forms the postero-lateral part of the roof of the skull and bears two 
dorsal crests. The lateral crest is the continuation of the crest on the dorsal surface 
of the sphenotic, whilst the more medial crest originates beneath the lateral edge of 
the frontal and extends postero-laterally to connect with the lateral crest about mid- 
way along the length of the pterotic. This medial crest forms the posterior edge of 
the incomplete roof to the post-temporal fossa, and the combined crests posteriorly 
form the lateral boundary of the unroofed portion of the post-temporal fossa. 

The supraorbital canal passed forwards within the frontal giving rise to several 
subsidiary branches which passed on to the surface of the frontal. Two branches 
passed postero-medially from the region of the centre of ossification, a further branch 
passed antero-medially and opened on to the interorbital area through a slit-like 
opening. The main supraorbital canal continued anteriorly within the frontal. 
The infraorbital sensory canal appears to have passed through part of the dorsal sur- 
face of the sphenotic and opened on to the surface medial to the sphenotic crest. The 
otic branch of the infraorbital sensory canal entered the pterotic crest postero-lateral 
to the rear end of the post-temporal fossa and passed forwards within the crest. The 
tube which contained the sensory canal opens anteriorly on the surface of the pterotic 
in the angle between the two pterotic crests. The contained sensory canal passed 
forwards on the surface of the pterotic and the sphenotic to connect with the main 
infraorbital canal on the dorsal surface of the latter bone. A supratemporal com- 
missure does not appear to have been present. 

On the edge of the frontal, in the anterior orbital region, a small supraorbital bone 
is present. The supraorbital bears an external ornamentation of raised ridges of 
bone radiating ventrally from a point near to the mid-dorsal border. 

The mesethmoid is prominent but not heavily ossified. It extends from the an- 
terior end of the frontals and ends in an acute point. Its basic form is that of two 
laminae of bone fused irregularly in the anterior region and diverging posteriorly to 
lie alongside the antero-lateral edges of the frontals. The two backwardly divergent 



ESPECIALLY MYCTOPHOIDS 



4i 



processes are slightly rolled at the edges to give a longitudinally concave dorsal 
aspect. Woodward (1902 : 45, pi. 12, fig. 1) termed these pieces the nasals, due to 
their elongated and rolled form. It is quite possible that fusion has occurred and 
that the nasals have become combined with the mesethmoid, no trace of nasals 
having been found in the specimen. The lateral edges of the mesethmoid lie along 




spo 



Fig. 17. Cimolichthys levesiensis Leidy. Neurocranium in ventral view. 
From B.M.N. H. number P.1811. 



4 2 



UPPER CRETACEOUS TELEOSTS 



the dorso-medial regions of each palatine thus providing an articulatory facet for 
those bones. 

The vomer is large and wide although extremely shallow. Anteriorly it tapers to a 
point and is attached to the underside of the extreme anterior end of the meseth- 
moid. The vomer widens posteriorly and its edges are in close association with 
longitudinal ventro-medial facets on the palatines. More posteriorly below the 
lateral ethmoids the vomer tapers and merges into the parasphenoid. The vomer is 



spo 




Fig. i 8. Cimolichthys levesiensis Leidy. Neurocranium in lateral view. The inset 
shows a diagrammatic representation of the prootic in antero-lateral view. From 
B.M.N. H. number P.1811. 



toothed in its anterior region, and the teeth are in a median row, there being four 
teeth in all. The bases of the teeth are clearly observable although none of the teeth 
were present on the specimen examined. 

The lateral ethmoids are poorly ossified bones, especially medially. Each is com- 
posed of thin sheets of bone which represent superficial ossifications of cartilaginous 
blocks. Dorsally, each lateral ethmoid is attached to the underside of the frontals, 
and ventrally the only well ossified part of the bone occurs. This is represented by a 
small crenulated knob, in life capped by cartilage, which articulates with a definite 
area at the posterior end of the dorsal region of each palatine. 

The parasphenoid is long and bent through a shallow angle at the base of the 
ascending process. The anterior region of the bone is flattened and expanded 
where it contacts the vomer. The parasphenoid narrows below the orbit and attains 
its minimum width at the base of the ascending process. The ascending process 
passes dorsally and contacts the prootic. The foramen through which the internal 
carotid artery passed into the base of the orbit lies in the ventral region of the ascend- 
ing process. Below this foramen on the ventro-lateral surface of the parasphenoid 
there is a shallow depression with which the pharyngobranchial of the first branchial 
arch articulated. Posteriorly the parasphenoid just contacts the ventral edge of the 
exoccipital and attaches to the ventro-lateral edge of the basioccipital. The para- 
sphenoid ends posteriorly below the hind end of the basioccipital leaving a small 



ESPECIALLY MYCTOPHOIDS 43 

fenestra between it and the basioccipital. This fenestra opens into the myodome 
posteriorly, the parasphenoid forming the entire floor of the posterior myodome. 

The basioccipital forms the whole of the occipital condyle. The condyle is circular 
and concave with a very small notochordal pit in its centre. Anteriorly the basi- 
occipital narrows and its undersurface passes antero-dorsally above the myodome 
forming part of its roof. As well as the posterior part of the roof of the myodome, 
the basioccipital also enters into the dorso-lateral parts of the wall of the myodome. 
On the lateral face of the basioccipital near to the condyle there is a group of pits and 
ridges. Internally the basioccipital forms the walls of the posterior parts of the 
otolith chambers, and as the ventral surface of the bone is inclined antero-dorsally 
above the myodome the otolith chambers become partially separated in the mid-line. 
The basioccipital contacts the rear edge of the prootic bridge anteriorly, and the 
otolith chambers extend forwards as narrow recesses within the prootics below the 
facial formen. 

The prootic is a large and extremely complex bone. Each prootic consists of two 
vertical sheets of bone joined anteriorly and ventrally. The inner sheet contacts the 
basioccipital behind, separating the myodome from the otolith chamber, whilst the 
outer sheet forms the wall of the otolith chamber, meeting the parasphenoid below 
and the exoccipital posteriorly. Dorsally the inner sheet curves medially meeting its 
fellow in the mid-line, forming the prootic bridge which separates the myodome 
from the cranial cavity. The lateral sheet curves dorso-laterally to contact the 
sphenotic and pterotic in the hyomandibular facet and the exoccipital posteriorly. 
These two divergent sheets of bone are connected anteriorly by a sheet of bone lying 
in the postero-lateral face of the orbit. This meets the sphenotic dorso-laterally and 
the pleurosphenoid dorso-medially, and its medial edge forms the margin of the optic 
fenestra. The trigemino-facialis chamber, within the prootic, is divided by a bony 
partition into a medial pars ganglionaris and a lateral pars jugularis. The pars 
jugularis is a long horizontal canal, widest anteriorly, lying within the prootic 
lateral to the prootic bridge and in the angle between the dorsal and ventral parts of 
the lateral face of the bone. Two foramina lead from the cranial cavity into the pars 
jugularis, the more posterior one being the facial foramen which transmitted the 
hyomandibular and palatine branches of the facial nerve. The anterior, trigeminal 
foramen transmitted the trigeminal and the buccal, otic and superficial ophthalmic 
branches of the facial nerve. The pars jugularis has four external openings. The 
posterior one is an oval foramen opening on the hind edge of the prootic, and this 
transmitted the jugular vein. Dorso-laterally, below the anterior region of the hyo- 
mandibular facet there is a second large dorso-ventrally elongated foramen which 
transmitted the hyomandibular branch of the facial nerve outwards from the pars 
jugularis. Below the jugular canal there is a third foramen which opens into a canal 
passing up into the floor of the pars jugularis, this canal transmitted the orbital 
artery into the jugular canal. The fourth foramen is the largest and is the anterior 
opening of the pars jugularis in the hind wall of the orbit. The trigeminal nerve, the 
buccal and otic branches of the facial nerve, the superficial ophthalmic components 
of both nerves, the jugular vein and the orbital artery all passed through this anterior 



44 UPPER CRETACEOUS TELEOSTS 

opening. The otic nerve passed dorso-laterally on to the undersurface of the sphen- 
otic to enter the bone through a small foramen. The superficial ophthalmic nerves 
passed antero-dorsally within a groove continuous across the face of the pleuro- 
sphenoid to the ventral surface of the frontal. The profundus nerve does not appear 
to have entered the pars jugularis, there being a small foramen in the hind wall of the 
orbit medial to the anterior opening of the pars jugularis which may have transmitted 
it. The oculomotor nerve passed out of the cranial cavity through the optic fenestra 
and its passage is marked by a deep notch on the medial edge of the prootic just above 
the prootic bridge. The palatine nerve, which entered the pars jugularis through the 
facial foramen, passed out through a canal in the floor of the pars jugularis. This 
palatine canal is confluent with the canal transmitting the orbital artery up into the 
pars jugularis. Within the myodome, part of the internal wall of the canal is absent 
so that the confluence of the two canals is clearly shown with that of the palatine 
nerve entering the myodome and that of the orbital artery entering from the lateral 
face of the prootic. Two further very small foramina are found in the lateral wall of 
the prootic dorsal to the foramen of the orbital artery. One of these opens into the 
same canal which housed the orbital artery and palatine nerve, and could possibly 
have transmitted a subsidiary branch of the palatine nerve. More likely, however, 
the two foramina transmitted a small artery and vein supplying the muscle masses in 
the area above the prootic. The foramen for the abducens nerve is postero-medial to 
the facial foramen, and the nerve passed ventrally into the myodome. 

The exoccipitals meet in the mid-line of the posterior face of the neurocranium both 
above and below the wide, shallow foramen magnum, completely enclosing it. It has 
already been noted that the exoccipitals do not enter into the composition of the 
occipital condyle but each exoccipital is considerably thickened at the lower corners 
of the foramen magnum. Several small foramina are found in the exoccipital on 
either side of the foramen magnum and these probably transmitted occipital nerves to 




f.hm 



10mm 



par 



Fig. 19. Cimolichthys levesiensis Leidy. Neurocranium in posterior view. 
From B.M.N. H. number P.1811. 



ESPECIALLY MYCTOPHOIDS 45 

the vertebral column. A slight dorso-laterally inclined ridge is produced on the 
exoccipital and is continued upwards by the epiotic as a prominent strut. The 
exoccipital also appears on the lateral face of the neurocranium since the bone is 
angled mesially, part lying on the posterior face of the neurocranium and part on the 
lateral face. The intercalar is attached to the outside of this angle. Laterally the 
exoccipital contacts the basioccipital ventrally, the prootic anteriorly and the 
pterotic dorsally. In the postero-ventral angle of the exoccipital the foramen which 
transmitted the vagus nerve is seen. This foramen is obscured in lateral aspect by a 
downwardly directed flange of the exoccipital. Slightly anterior to the vagus fora- 
men is the much smaller foramen for the glossopharyngeal nerve. 

The intercalar is a small flattened bone on the hind end of the lateral face of the 
neurocranium, elongated dorso-ventrally and ending below the posterior end of the 
hyomandibular facet. It does not take part in the composition of the wall of the 
cranial cavity, being entirely superficial in position and simply providing a point of 
articulation for the ventral limb of the large post-temporal bone. 

The pterotic has already been described in dorsal aspect ; laterally it forms the 
major part of the hyomandibular facet and is excavated above the facet to produce 
the posterior region of the dilatator fossa which is a prominent oval fenestra. The 
pterotic contacts the sphenotic and prootic in the anterior region of the hyomandibu- 
lar facet. 

The sphenotic is large and forms the postero-dorsal corner of the orbital region, the 
antero-dorsal part of the hyomandibular facet and the anterior half of the dilatator 
fossa. The hyomandibular facet is shallow posteriorly, anteriorly however it is a 
well defined cup-shaped facet formed dorsally of sphenotic and ventrally of prootic. 
Within the orbit the sphenotic contacts the prootic and the pleurosphenoid, and 
bears a small dorso-laterally directed foramen which transmitted the otic nerve into the 
sphenotic to innervate the infraorbital sensory canal. 

The pleurosphenoids, in the hind wall of the orbit, meet the frontals, sphenotics and 
prootics. The dorsal region of each pleurosphenoid is only ossified superficially and 
the more medial regions do not contact the frontal, the contact having been effected 
through the intervention of cartilage. The groove on the face of the bone, already 
mentioned in connection with the superficial ophthalmic nerves, passes up towards 
the ventral surface of the frontal. The course of the nerves along the frontal is in- 
dicated by a row of small foramina through which branches of the nerve passed to 
the supraorbital sensory canal. 

Hyopalatine bones. The hyopalatine bones are shown in medial view in Text- 
figure 20. The hyomandibular is short, broad and vertical in position, the quadrate 
condyle being immediately below the centre of the hyomandibular head. The head 
region is elongated, and inclined antero-ventrally. The anterior part of the head is 
thickened and associated with the anterior cup-shaped region of the hyomandibular 
facet. The remainder of the head is straight and narrow corresponding with the 
pterotic portion of the facet. The opercular process is not well defined and is merely 
a thickening of the posterior edge about mid-way down the bone. On the lateral 
face of the hyomandibular a crest runs postero-ventrally from the anterior region of 



4 6 



UPPER CRETACEOUS TELEOSTS 
enp t 




sop 



Fig. 20. Cimolichthys levesiensis Leidy. Hyopalatine and opercular bones and 
mandible of the right side in medial view. From B.M.N.H. P.1811. 

the head becoming lower and finally terminating near the ventral end of the bone. 
The preoperculum lies against the posterior face of this crest and is overlapped 
slightly by it. The hyomandibular tapers ventrally and its anterior extent is thin 
and covered laterally by the metapterygoid. The hyomandibular branch of the 
facial nerve entered the bone through a large foramen in the antero-dorsal region of 
the medial face immediately posterior to the anterior hyomandibular head. The 
nerve passed within the bone following a course parallel with that of the crest on the 
lateral face, and divided to give rise to the three main branches of the hyomandibular 
nerve. The opercular nerve left through a foramen above the opercular condyle in 
the angle between the main body of the hyomandibular and the lateral crest. The 
mandibular nerve passed through a foramen below the opercular condyle and here 
the posterior edge of the hyomandibular is excavated slightly. The mandibular 
nerve continued ventrally along a groove in the posterior edge of the hyomandibular. 
The hyoidean nerve may have left in the same region as the mandibular nerve since 
there is a second smaller foramen below the condyle. 

On the internal face of the quadrate a short stout bone is inclined forwards ven- 
trally and is slightly mesially constricted. This bone by its position suggests a 
symplectic, but in shape it resembles an interhyal. 

The quadrate is triangular with a prominent transverse condyle. The groove 
which housed the symplectic ends ventrally above the condylar region. The post- 
erior edge of the quadrate is slightly thickened and lies against the thickened edge 
of the lower part of the preoperculum. Anteriorly the quadrate overlaps, laterally, 
the posterior extent of the ectopterygoid and is firmly attached to it. 



ESPECIALLY MYCTOPHOIDS 47 

The ectopterygoid is bent through a slight angle towards its posterior end where it 
slopes postero-ventrally to meet the quadrate. Anteriorly the ectopterygoid con- 
tacts the postero-ventral region of the palatine and there is a slight angle between the 
two bones, the palatine not forming a direct continuation of the ectopterygoid. The 
ectopterygoid bears a single row of teeth ventrally which are not regularly placed. 
The teeth are firmly fused to the bones and have large expanded bases which are 
practically circular in cross-section. The teeth themselves become laterally com- 
pressed apically to give rise to anterior and posterior cutting edges. The posterior 
cutting edge is further increased by the production of a post-apical barb. The teeth 
decrease in size posteriorly. The internal edge of the bone is inclined dorso-medially 
where it contacts the endopterygoid. 

The endopterygoid is a thin sheet of bone lying in a horizontal plane and its in- 
ternal edge is associated with the lateral edge of the parasphenoid forming the roof of 
the mouth. Dorsally the endopterygoid is concave where it lies below the orbit. 
The convex ventral surface is covered by a large, roughly oval patch of minute teeth. 
Posteriorly the endopterygoid lies medial to the anteriormost extension of the meta- 
pterygoid. 

The metapterygoid is a thin laminate bone superficially placed so as to cover the 
posterior extent of the endopterygoid and the anterior expansion of the hyomandibu- 
lar. Posteriorly the metapterygoid is irregularly rounded but anteriorly there is an 
oblique ridge forming the upper part of the anterior border; the lower part of the 
anterior border expands slightly in front of this oblique strengthening ridge. 
Ventrally the metapterygoid meets the ectopterygoid and the dorsal edge of the 
quadrate. 

The palatine is large and elongated. The ectopterygoid attaches to its ventral 
surface near the posterior end, which is expanded towards the mid-line, irregularly 
ridged dorsally and was in life filled by cartilage, providing the articulatory facet for 
the ventral region of the lateral ethmoid. The palatine has a convex dorsal surface, 
but the ventral surface is concave so that an elongated ventral groove is present 
longitudinally. Anteriorly the palatine tapers to a point and ends on a level with 
the anterior mesethmoid region. Along the medial face of the palatine two longi- 
tudinal grooves are present, a dorso-medial groove along the anterior half of the bone 
with which the lateral edge of the mesethmoid articulated, and a ventro-medial 
groove extending practically the whole length of the bone into which the lateral edge 
of the expanded vomer fits. On the lateral face of the palatine near to the posterior 
end there is a shallow obliquely orientated groove which houses the anterior end of 
the maxilla. In the ventral longitudinal groove of the palatine two major tooth 
rows are found (Woodward, 1902 : 45, fig. 10). The inner row is the more im- 
portant and is a direct continuation of the row of ectopterygoid teeth. Anteriorly 
the bone narrows and the teeth are reduced to a single row. At the posterior end of 
the palatine the teeth become much enlarged and the bone ends with one enormous 
tooth of the inner row. The teeth are identical to those on the ectopterygoid except 
that the outer teeth are not provided with a post-apical barb. The inner row of 
teeth is represented by 11 tooth sockets or teeth and the teeth in the middle of the 



<8 



UPPER CRETACEOUS TELEOSTS 



row are the smallest. The n sockets would presumably indicate the presence of 
about 5 or 6 functional teeth together with their replacements. The posterior 
tooth of the palatine and the anteriormost tooth of the ectopterygoid are equal in 
size and are the largest teeth in the upper jaw. The teeth of the inner row are 
closely arranged, and their bases are ovoid in shape. The outer row of teeth is 
of less importance and is represented by smaller unbarbed teeth, which, unlike 
the inner row, reach their maximum size half way along the bone and decrease in 
size anteriorly and posteriorly. There are 8 tooth bases present, again probably 
functional teeth and their successors. Unlike the inner row, the sockets are not 
quite so crowded together and have a more circular outline. Loomis (1901) has con- 
sidered the histological structure of the teeth of certain species of this genus from 
the Niobrara Chalk of the United States. 




sop 



art 



Fig. 21. Cimolichthys levesiensis Leidy. Restoration of the skull in lateral view. 



Dermal upper jaw. The dermal upper jaw is shown in lateral view in Text-figure 
21. The jaw consists of a premaxilla, a maxilla and a small supramaxilla. The 
premaxilla is a large, antero-posteriorly elongated lamina of bone, tapering at 
both ends. It is more than four times longer than its maximum depth and its lateral 
face is ornamented with radiating rows of small bony tubercles. The tubercles 
radiate ventrally from a point near to the dorsal border of the premaxilla at the region 
of its maximum depth which is just within the anterior half of the bone. Anteriorly 
the premaxilla curves around the acutely pointed snout and probably contacted its 
partner in the mid-line. The oral border of the premaxilla bears a single row of 
small, conical, closely arranged teeth. 



ESPECIALLY MYCTOPHOIDS 49 

The maxilla is an elongated, stout bar of bone which is straight for most of its 
length except anteriorly where it is curved medially. The maxilla is smooth, un- 
ornamented and bears no traces of having been toothed. The premaxilla forms the 
anterior half of the oral border of the mouth and the maxilla enters the gape behind 
it. Anteriorly the maxilla lies in the oblique groove on the lateral face of the pala- 
tine, and the curved head of the maxilla is inflected towards the mid-line above the 
dorso-lateral surface of the palatine. The maxilla was attached to the premaxilla 
ligamentously. 

A small supramaxilla is present on the dorsal edge of the maxilla near to its hind 
end. The supramaxilla is elongated antero-posteriorly, markedly curved and orna- 
mented slightly with tuberculated ridges. 

Mandible. The mandible is shown in medial and lateral views in Text-figures 20 
and 21. It is long and relatively shallow, its maximum depth being less than one- 
fifth of its length, although there is a considerable horizontal extent to the mandible 
where the ventral edge is curved medially. Anteriorly the mandible tapers to the 
constricted symphysis. The dentary forms much of the mandible with upper and 
lower thickened regions joining anteriorly. The thickened dorsal limb bears 3 rows 
of teeth on its oral side. The extreme oral border bears a close series of minute 
teeth. Medial to this row there is a second row of larger teeth which are irregularly 
distributed and unevenly spaced. The internal teeth are the largest and there 
are approximately 7 in the row. The largest tooth is the second and tooth size 
decreases posteriorly. The teeth do not bear the post-apical barb so characteristic 
of the palato-pterygoid teeth, however they are laterally compressed with anterior 
and posterior cutting edges. All of the teeth in the three rows are conical, hollow, 
and firmly fused to the jaw by an unexpanded base. The largest teeth are more 
markedly recurved than the smaller ones. 

The articular forms the posterior mandibular region and is inserted on to the 
internal face of the dentary and occludes the posterior V-shaped indentation of the 
latter bone. The articular facet is shallow with a mesial ridge corresponding to the 
constricted condyle. There is no retroarticular process behind the articular facet so 
that the facet is vaguely defined. Beneath the facet, on the lateral face of the 
articular, traces of a groove are present in which the mandibular sensory canal ran. 
The canal was continued in a ventro-lateral groove along the surface of both the 
articular and the dentary. Ventrally the mandible is considerably thickened and 
ornamented by prominent longitudinal bony ridges extending forwards from 
beneath the articular facet, and backwards from the symphysial region on the 
dentary. The remainder of the lateral face of the mandible is smooth except for the 
oral border of the dentary where a narrow band of longitudinal ridging is observed. 

A minute angular element is visible on the extreme postero-ventral angle of the 
mandible. 

Body. The vertebral column is estimated to consist of approximately 60 verte- 
brae (50 are shown in B.M.N.H. specimen number 38113). Of these approxi- 
mately 30-32 are caudal. The precaudal vertebrae bear long narrow ribs supported 
on small transverse processes. The ribs extend laterally around the abdominal 



50 UPPER CRETACEOUS TELEOSTS 

cavity. All of the vertebrae bear neural arches and spines dorsally, those spines in 
the precaudal region being long and very slender. In the caudal region the neural 
spines become shorter and much stouter. The centra are slightly longer than deep 
and are mesially constricted. Each centrum is strengthened by one or two lateral 
longitudinal ridges. Epineurals and epipleurals occur on the precaudal vertebrae. 
The pectoral fins are low down on the flanks and each fin appears to consist of 
approximately 12 rays. The rays are stout and well ossified with the first ray 
unbranched. The pelvic fins are abdominal, situated below the mid-point of the 
dorsal fin. Each pelvic fin consists of 10 or 11 rays. The pelvic bones are large 
triangular plates meeting in the mid-line of the ventral body wall. The pelvic 
fins are almost as large as the pectoral fins. The dorsal fin lies within the anterior 
half of the trunk and is relatively short based. It is composed of approximately 
15 rays, but more may have been present. The anal fin is only shown by a few 
imperfect fragments, and all that can be said is that possibly as many as 10 rays 
were present ; it is remote in position, quite close to the caudal fin. The caudal 
fin is not represented but part of the caudal skeleton can be seen. The preural 
vertebrae are about four in number and the centra are reduced in length. The 
neural and haemal spines are enlarged and somewhat flattened. The first preural 
vertebra is fused with the first ural vertebra, and the compound centrum so pro- 
duced supports the parhypural (the haemal spine of the first preural vertebra) and 
the first two hypurals. These hypurals are fused into a single compound element. 

Scales. The trunk is devoid of normal squamation but bears two rows of scutes 
on each flank, one placed dorso-laterally, the other ventro-laterally. The scutes 
(Woodward 1902, pi. 12, fig. 5) are overlapping and ' arrowhead ' shaped, those in 
the anterior region of the dorso-lateral row being the largest. Each scute bears a 
longitudinal raised keel, and the posterior region is broader and slightly corrugated. 
The scales are longer than they are broad with the anterior region narrow, smooth 
and slightly lengthened. At the base of the pelvic fin, one of the scutes in the 
ventro-lateral row has its lower margin excavated to allow for the fin insertion. 



Family DERCETIDAE Pictet, 1850 

Diagnosis (emended). Body elongated and very shallow ; head elongated to a 
greater or lesser extent. Post-temporal fossa never completely roofed. Maxilla 
enters gape and supports a single tooth row. Supraorbital and supramaxilla absent. 
Vertebrae 60-80 in number of which approximately 30 are caudal. Anterior 
precaudal vertebrae often elongated. Ribs borne on prominent laminar transverse 
processes. Neural and haemal spines always short. Intermuscular bones extremely 
abundant. Scales normally confined to one or more continuous rows of enlarged 
triradiate scutes along the flank, occasionally the remainder of the body covered by 
much smaller scales. 



ESPECIALLY MYCTOPHOIDS 51 

Genus DERCETIS Agassiz, 1834 

1863 Leptotrachelus von der Marck: 59. 

1940 Benthesikyme White and Moy-Thomas : 102. 

Diagnosis (emended). Dercetidae in which the head is more or less extended, 
with premaxilla produced slightly in advance of the mesethmoid. Body region 
always extended. Mandible extends to the anterior tip of the snout. Teeth 
clustered and in the form of slender hollow cones, present on premaxillae, maxillae, 
dentaries, palatines and ectopterygoids. Paired fins prominent with pectorals larger 
than the pelvics. Dorsal fin more or less extended, in the mid-region of the back. 
Anal fin short and remote. A single pair of transverse processes per centrum. 
Anterior precaudal centra at least twice as long as deep. Dermal scutes in a paired 
series along dorsal and ventral borders and in a single series along lateral line ; a 
few large lath-shaped intermediate scutes in larger species, but squamation never 
complete. 

Type species. Dercetis scutatus Agassiz. 

Remarks. The genus was erected by Agassiz (1834 : 39°> an( i Feuill. ; 20) to 
include the type species Dercetis scutatus from the Upper Senonian deposits of 
Baumberg, Westphalia. Later von der Marck (1863) erected another genus within 
the family Dercetidae to include the new species Leptotrachelus armatus (59, pi. 10, 
fig. 3). In 1873, von der Marck added a further species, Leptotrachelus sagittatus 
(63, pi. 2, fig. 1). Both of these species were erected on imperfect material from the 
Upper Senonian (Campanian) of Sendenhorst, Westphalia, and his type material is 
located in the Palaontologisches Institut der Westfalische Wilhelms-Universitat, 
Munster. Since 1873 several more species of both genera, Dercetis and Lepto- 
trachelus, have been erected, the material coming from various localities such as the 
Chalk of S.E. England, the Niobrara Chalk of Kansas and from the Sahel Alma 
locality in the Lebanon. 

The holotype of Dercetis scutatus (at one time in the Bayerische Staatssammlung fur 
Palaontologie, Munich) appears to have been either lost or destroyed. Unfor- 
tunately no other specimens of the type species could be traced. After examination 
of most of the species placed in the genus Dercetis from the English Chalk, together 
with the examination of both the Leptotrachelus species erected by von der Marck, 
and those leptotrachelids from the Lebanon, it is clear that they all belong in a single 
genus. This conclusion is in accordance with that of Siegfried (1966 : 215), who 
has examined and redescribed von der Marck's material, and as a result moved both 
to the genus Dercetis (Dercetis armatus (Text-fig. 24) and Dercetis sagittatus). 

Several species of Leptotrachelus from Sahel Alma have been studied but all of the 
material is fragmentary and badly preserved. The descriptions are therefore 
incomplete but even so are sufficient to show that these species also belong to the 
genus Dercetis. The three species from Sahel Alma, Dercetis triqueter (Text-fig. 23), 
Dercetis gracilis (Text-fig. 25) and Dercetis rostralis (Text-fig. 26), although still 
considered to be three separate species, are arbitrarily separated. Thus considering 
all the Dercetis material available a graded series is apparent, ranging from the 



5-' UPPER CRETACEOUS TELEOSTS 

largest Dercetis species, Dercetis maximus from the English Chalk (Woodward, 
1903, pi. 15, fig. 2), to the smallest Dercetis gracilis from the Lebanese chalk. 

White and Moy-Thomas (1940 : 102) pointed out that the generic name Lepto- 
trachelus was preoccupied and erected a new generic name, Benthesikyme. Since 
all of the species of Leptotrachehis are now included in the genus Dercetis, Ben- 
thesikyme is treated as a synonym. 

Dercetis triqueter Pictet 
(Text-figs. 22, 23) 

1850 Dercetis triqueter Pictet : 47, pi. 9, figs. 5, 6. 
1850 Dercetis linguifer Pictet : 47, pi. 9, figs. 7, 8. 
1866 Leptotrachelns triqueter (Pictet) Pictet & Humbert : 95, pi. 14, figs. 1, 2. 



Davis : 620, pi. 38, fig. 1. 

Woodward : pi. 10. 

Woodward : 174. 

Woodward : 68, text-figs. 15, 16. 

Woodward : fig. 3. 

Berg : fig. 215. 

Signeux : 642, pi. fig. 4. 

Bertin & Arambourg : fig. 161 7A. 



1887 Leptotrachehis triqueter (Pictet) 

1898 Leptotrachelus triqueter (Pictet) 

1901 Leptotrachelus triqueter (Pictet) 

1903 Leptotrachelus triqueter (Pictet) 

1926 Leptotrachelus triqueter (Pictet) 

1940 Leptotrachelus triqueter (Pictet) 

1954 Leptotrachelus triqueter (Pictet) 

1958 Leptotrachelus triqueter (Pictet) 

1964 Benthesikyme triqueter (Pictet) Danil'chenko : fig. 141. 

1966 Leptotrachelus triqueter (Pictet) ; Lehman : fig. 200. 

Diagnosis (emended). Dercetis of standard length up to 22-5 cm. Vertebrae 
between 70 and 73 in number of which 30 are caudal. Anterior precaudal centra 
twice as long as deep. Head equal to approximately one-fifth of the standard 
length and the maximum height of the head equal to one-third of its length. Pre- 
orbital region forms 60% of the total head length. Mesethmoid bluntly bifurcate 
anteriorly. Dorsal fin with 35 rays ; anal fin with 8 or 9 small feeble rays, remote in 
position. Pectoral fin with 11 rays ; pelvic with 8 rays arising behind anterior third 
of the trunk. Scutes of varying sizes, with the precaudal ones heart-shaped ; 
triradiate pattern obscured in many. Intermediate lath-shaped scutes often 
present. 

Holotype. Fragment of abdominal region in the Museum d'Histoire Naturelle, 
Geneva, from the Upper Senonian of Sahel Alma, Lebanon. 

Material. Specimens in the B.M.N.H., the Museum d'Histoire Naturelle in 
Geneva, and the Museum National d'Histoire Naturelle, Paris. 

Remarks. The species was originally erected by Pictet as Dercetis triqueter and 
in the same work (Pictet, 1850) he erected a second species, D. linguifer. Pictet 
and Humbert (1866 : 95) synonymised both species under Leptotrachelus triqueter. 
Both holotypes {Dercetis triqueter and Dercetis linguifer) are in Geneva. That of 
Dercetis linguifer is also a small fragment of an abdominal region and is clearly 
referable to Dercetis triqueter. 

Description. Neurocranium. The neurocranium is indistinct and badly 
preserved in all of the specimens examined. The neurocranium is elongated, 



ESPECIALLY MYCTOPHOIDS 53 

shallow and widest at the hind end of the orbits. The frontals make up practically 
all of the skull-roof reaching back almost to the occipital border. Anteriorly the 
frontals taper to contact the rear edge of the mesethmoid near to the front end of 
the snout. The frontals widen to form the upper margin of the orbits. Posteriorly 
the frontals contact the supraoccipital and the parietals. The ornamentation on 
the frontals radiates in all directions from the centre of ossification above the rear 
end of the orbit. Rows of minute tubercles pass anteriorly and antero-laterally 
above the orbit, and posteriorly and postero-medially above the cranial cavity itself. 
Towards the lateral edge of the frontal above the orbit the ornamentation is no 
longer in uniform rows but is randomly scattered. The supraoccipital is small, 
meeting the frontals anteriorly and forming the centre of the occipital border 
of the skull-roof. The parietals lie lateral to the supraoccipital and are small 
transversely orientated bones ornamented with tubercles. The pterotic forms the 
postero-lateral border of the skull-roof, and the rows of tubercles on its dorsal 
surface originate near the hind end of the bone. The sphenotic projects laterally 
from beneath the frontal at the rear end of the orbit, but its dorsal surface appears 
to have been unornamented. 

The mesethmoid is in the form of two strips of bone joined anteriorly. These 
strips flank the extreme anterior ends of the frontals. The mesethmoid is slightly 
expanded anteriorly into two small lateral wings and bears a median, anterior, 
U-shaped indentation which imparts a slight bifid appearance. The vomer corres- 
ponds to the mesethmoid anteriorly in the possession of lateral wings. The wings 
are flattened and lie in a horizontal plane and fit into antero-posteriorly elongated 
facets on the antero-medial faces of the palatine. The vomer only meets the 
mesethmoid in this anterior region, widening behind to pass back as a prominent 
flattened plate which attaches to the parasphenoid. Teeth do not appear to have 
been present on the vomer. The parasphenoid is a straight rod of bone running 
through the base of the orbit, apparently without lateral flanges. 

Hyopalatine bones. The hyopalatine bones are shown in medial view in Text- 
figure 22. The jaws are elongate and bear a prominent armature of long teeth. 
The palatine bone is long and narrow, and in the form of a curved, thickened lamina 
with a longitudinal ventral concavity housing the teeth. In the extreme anterior 
region the bone narrows and the concavity is lost. The medial face of the palatine 
is excavated producing a facet for the reception of the lateral wing of the meseth- 
moid. Anteriorly the palatine ends in advance of the mesethmoid, and posteriorly 
below the anterior region of the orbit. The lateral ethmoid is not prominent and 
ventrally is weakly associated with the posterior end of the palatine. The teeth 
are closely packed and give the appearance of having been arranged in three major 
rows. The teeth are all of the same type, there being minor variations in length 
and breadth, but all are needle-like with unexpanded basal regions. The diameter 
of the tooth bases varies little between the large teeth and the small teeth, the major 
change in dimension is in the length. All of the teeth are hollow and oval in cross- 
section. The cutting edge of each tooth is confined to the extreme apex where the 
hollow cavity is occluded and the cross-section of the tooth becomes more elliptical. 



54 



UPPER CRETACEOUS TELEOSTS 



The cutting edge extends around the apex imparting a slight barbed appearance to 
the tooth both pre-apically and post-apically. The teeth are practically straight 
and most of them are inclined slightly anteriorly rather than vertically downwards. 
The teeth are shortest at the anterior end and increase in size posteriorly reaching 
their maximum length about two-thirds of the way back, after which they again 
decrease in size. The palatine contacts the ectopterygoid below the front end of 
the orbit and it appears that one of the rows, the most internal, of the palatine teeth 



hm.h 



op.p.hm 




art 



Fig. 22. Dercetis triqueter Pictet. Hyopalatine bones and mandible of the left side 

in medial view. 



is continued back on the ectopterygoid. The teeth on the ectopterygoid are, 
however, shorter, stouter and markedly recurved. Each tooth still exhibits the 
apical cutting region both in front of and behind the apex. The ectopterygoid is 
small in comparison to the palatine and inclined at a greater angle to the horizontal. 
Posteriorly the ectopterygoid meets the quadrate and overlies it laterally. The 
endopterygoid attaches to the internal region of the ectopterygoid and is in the form 
of a flat plate lying medially in the roof of the mouth. The posterior region of the 
ectopterygoid is thickened in two places, the thickenings diverging posteriorly. 
The lower thickened region contacts the quadrate, the upper thickened region passes 
postero-dorsall}' and forms the attachment region for the rear edge of the endo- 
pterygoid. This latter thickening is continued postero-dorsally by a thickened 
crest on the anterior region of the metapterygoid. This bifurcation of the thickened 
edges of the ectopterygoid is clearly visible from the medial aspect. The meta- 
pterygoid is large with an obliquely inclined crest, the remainder of the bone overlying 
the lateral face of the anterior region of the hyomandibular and attaching to the 
dorsal edge of the quadrate. The hyomandibular is upright with an elongated head 
and a shallow lateral crest passing down the rear edge of the bone. The quadrate 
condyle lies below the level of the occipital border of the neurocranium. 

Dermal upper jaw. Due to the fragmentary nature of the specimens it is 
impossible to ascertain to what extent, if any, the maxilla enters the oral border of 
the upper jaw. Both premaxilla and maxilla are elongated and lie against the 
lateral face of the palato-pterygoid. Anteriorly the premaxilla is wrapped around 



ESPECIALLY MYCTOPHOIDS 55 

the end of the palatine. Thus medial to the anterior end of the palatine the pre- 
maxilla extends backwards and appears to be associated with the anterior edge of 
the mesethmoid within the U-shaped indentation of the latter bone. The pre- 
maxilla extends posteriorly as a shallow lamina of bone bearing a single marginal 
row of teeth on its ventral edge. The teeth are small but recurved and have the 
same apical specialization as the palatine and ectopterygoid teeth. The maxilla 
appears to have entered the gape posteriorly and borne teeth along its ventral edge. 

Mandible. The mandible is shown in medial view in Text-figure 22. The 
mandible is long and shallow, although it has a considerable ventral extent lying in 
a horizontal plane. Anteriorly the mandible is acutely pointed and the symphysis 
is constricted. The dentary is the major component of the lower jaw forming the 
entire oral border as well as most of the lateral face of the mandible. Teeth are 
borne on the oral edge of the dentary and are numerous and clustered. All of the 
teeth are long and needle-like and similar to those on the ectopterygoid in being 
recurved. The articular facet is ill-defined, being somewhat less than a semicircle 
and with no retroarticular process posteriorly. The outer face of the mandible is 
smooth and bears no traces of ornamentation. 

Vertebral column. The vertebral column is shown in the restoration in Text- 
figure 23. It consists of approximately 70 vertebrae although the number is slightly 
variable, some specimens appearing to have up to 73, although none has less than 
70. Of these 70 or so vertebrae 30 are caudal. The centra are longest in the 
anterior precaudal region and decrease in length posteriorly, those of the caudal 
region being uniform in size with the length of each centrum just in excess of its 
depth. In the anteriormost precaudal region the centra are at least twice as 
long as deep. All of the centra are mesially constricted and in the precaudal 
region this is especially noticeable. The characteristic feature of the precaudal 
region is the possession of prominent laterally directed transverse processes on each 
centrum. On the first few precaudal centra the transverse processes are relatively 
short and broad based, and borne on the anterior region of each centrum. Pleural 
ribs appear to articulate with the undersurface of each transverse process. The 
size of the transverse processes is constant back to about the 20th precaudal vertebra 
where they begin to elongate and incline slightly forwards. Thus while the centra 
decrease in length the transverse processes increase so that the longest transverse 
process, on the 35th precaudal vertebra, is just longer than the centrum supporting 
it. The transverse processes rapidly decrease in length and incline ventrally to fuse 
together and produce a haemal arch by the 40th vertebrae. Neural arches occur on 
all the vertebrae and are long and low. The neural spines arising from the arches 
are abbreviated and insignificant anteriorly but become more prominent towards 
the caudal region. The neural spines are straight. The neural arches occupy the 
whole of the dorsal extent of the corresponding centra, the neural spines being 
produced from the posterior end of each arch. Haemal arches and spines charac- 
terise the caudal region. The haemal spines remain small and feeble but, unlike 
the neural spines, exhibit a marked posterior curvature. The pleural ribs which 
articulate with the transverse processes are long and thin and directed posteriorly, 



5& 



UPPER CRETACEOUS TELEOSTS 










CD 

A 















fc 



^i.xi 
" IS 

<D 

XI 

d 
o 
> 



s 



g 



6 



ESPECIALLY M YCTOPHOIDS 57 

curving only slightly towards the ventral region. Thus an elongated, extremely 
thin ' neck ' region is produced, only appearing to widen near the dorsal and pelvic 
fins. The region below the vertebral column was presumably occupied by a highly 
distensible stomach region. Intermuscular bones, both epineurals and epipleurals, 
are present along much of the body. 

Median fins and tail. The median fins are shown in the restoration, Text-figure 
23. The dorsal fin is elongate and composed of approximately 35 rays, occupying 
the mid-third of the trunk. Anteriorly the fin rays are well formed, segmented and 
branched, the more posterior fin rays are small and feeble but still branched and 
segmented. The proximal radials have a short shaft region with an expanded 
stellate head for the articulation of the fin rays. 

The anal fin is small and remote, occurring well behind the dorsal fin and quite 
close to the caudal fin. Eight or 9 small rays are visible, all branched and segmented. 

The caudal skeleton appears to consist of at least 3 free preural vertebrae with a 
distinct free ural vertebra (ural 1) and possibly a terminal half-centrum representing 
the remains of ural vertebra 2. Apart from these facts practically nothing can be made 
of the caudal skeleton. The fin itself consists of 19 principal caudal rays of which 17 
are branched. Accessory fin rays above and below the caudal fin number about 12. 

Scales. The body is devoid of normal squamation but there are 3 rows of enlarged 
basically triradiate scutes (Text-fig. 82A) along each flank. The two major rows of 
scutes positioned dorso-laterally and ventro-laterally are triradiate. Each scute 
has an elongated anterior extension and two divergent posterior extensions ; the 
angle between the posterior arms is about 60 °. The scales on the dorso-lateral part 
of the flank have a longer upper posterior arm which curves over on to the dorsal 
surface of the body approaching its fellow of the opposite side. Similarly in the 
ventro-lateral row the lower posterior arm is the longer and it too curves round 
towards the ventral surface. Intermediate, lath-shaped scutes occur in two distinct 
rows on the caudal region of the body, as extensions towards the mid-line of the 
flank from the shorter posterior arms of the major scutes. The third row of scutes 
which occurs along the whole length of the body is positioned on the course of the 
lateral line. These scutes are smaller, markedly less triradiate, elongate and heart- 
shaped. All of the scutes in the three major rows bear a small backwardly projecting 
spine arising from the angle between the two posterior arms. These spines are more 
pronounced and uncinate in the caudal region. 

A great variability in the shape of the scutes occurs within the genus Dercetis and 
even within the species Dercetis triqueter, the nature of the scutes appearing to be 
dependent on the size of the specimen. Thus in the larger specimens the triradiate 
nature of the scales becomes obscured due to the expansion of the scale between each 
pair of arms. This expansion of the scales imparts an almost complete covering to 
the flank, the three major rows now practically contacting each other. 

Remarks. Two further species of Dercetis occur in the Sahel Alma deposits, 
Dercetis gracilis (Davis) (Text-figure 25) and Dercetis rostralis (Signeux) (Text-figure 
26) . The former is generally more fragile and smaller in size than the latter and the 



5* 



UPPER CRETACEOUS TELEOSTS 










IV 



0> 



<v 



<V 



u 

o 



V 

« 



> 

a 

Q 



3 

a 

fee 
on 



ac 



O 



6 



ESPECIALLY MYCTOPHOIDS 59 

total vertebral count is higher (80). The scale pattern would seem to be identical to 
that of the smaller specimens of Dercetis triqueter in which no expansion of the scales 
has occurred. Dercetis rostralis (Signeux) is a much more slender species than even 
Dercetis gracilis. The precaudal region has vertebrae at least six times as long as 
deep. The scute arrangement in this ' neck ' region is also simplified, with only 
a single row of scales (lateral line?), whereas in the caudal region the more typical 
squamation of the genus is exhibited, i.e. three rows of scales. 

Genus RHYNCHODERCETIS Arambourg, 1944 

Diagnosis (emended). Dercetidae in which the rostral region of the neuro- 
cranium is excessively elongated. Rostrum formed distally of premaxillae united 
in the mid-line. Mandible does not reach the anterior end of the premaxillae. 
Maxillae long, narrow and toothed, forming half of the gape. Teeth on dentaries, 
maxillae, palatines and ectopterygoids, but not on premaxillae. Paired fins reduced 
in size, pectoral fins on mid-flank region, pelvics arise behind the origin of the 
dorsal fin. Dorsal fin always short and midway along the body. Anal fin always 
short and remote. Two pairs of transverse processes on each precaudal centrum, 
the ribs articulating with the anterior pair. Dermal scutes reduced to a single row 
on each flank along the lateral line. 

Type-species. Rhynchodercetis yovanovitchi Arambourg. 

Remarks. The genus was erected by Arambourg (1944) to contain the single 
species Rhynchodercetis yovanovitchi, which is abundant in the Lower Cenomanian 
deposits of Morocco. Arambourg (1954) later expanded his original descriptions 
using new and more complete material from Morocco. Rhynchodercetis material 
from other localities is not abundant : Rhynchodercetis yovanovitchi has been 
described from Sicily by Leonardi (1966) ; Rhynchodercetis gortanii from Comen by 
d'Erasmo (1946) ; and Rhynchodercetis hakelensis from Hakel by Pictet & 
Humbert (1866). 

Rhynchodercetis hakelensis (Pictet & Humbert) 
(Text-figs. 27, 28) 

1866 Leptotrachelns hakelensis Pictet & Humbert : 98, pi. 14, fig. 3. 
1901 Leptotrachehts hakelensis Pictet & Humbert ; Woodward : 184. 

Diagnosis (emended). Rhynchodercetis of up to 10 cm. in standard length. 
Vertebral column with approximately 60 vertebrae of which 30 are caudal. Head 
length equal to one-quarter of the standard length. Maximum height of the head 
equal to one-fifth of the total head length. Preorbital length of the head forming 
70% of the total. Dorsal fin with 7 rays, anal fin with 8 rays. 

Holotype. Minute, incomplete fish in the Museum d'Histoire Naturelle, 
Geneva, from the Middle Cenomanian, Hakel, Lebanon. 

Material. One specimen, number P. 6001 in the B.M.N.H., prepared by the 
transfer method. The specimen is poorly preserved as are the other minute speci- 
mens in the B.M.N.H. 



6o 



UPPER CRETACEOUS TELEOSTS 



Description. Neurocranium. The neurocranium is shown in dorsal view in 
Text-figure 27. It is elongate and relatively shallow, and with an acutely pointed, 
elongated anterior region. The widest point of the skull-roof is immediately 
posterior to the orbits and is represented by the lateral extension of the sphenotics. 
The major part of the skull-roof is composed of frontals which extend almost back 
to the occipital border. The bones are relatively smooth, only bearing a few weak 
ridges which indicate the course of the sensory canals. The presence or absence of 
a post-temporal fossa is difficult to determine, but in the postero-lateral region of 
the neurocranial roof there is a definite groove which may represent an unroofed 
post-temporal fossa. More anteriorly, however, the frontal is flat and perhaps 
indicates that if the fossa extended into this region then it was roofed. The frontals 
meet in the mid-line and extend backwards to cover the anterior regions of both the 
supraoccipital and the parietals. The supraoccipital thus separates the parietals in 
the mid-line. The supraoccipital bears a prominent posteriorly directed crest 
which does not rise above the plane of the skull-roof. The frontals form the upper 
margins of the orbits and taper anteriorly to insert on to the posterior edges of the 
mesethmoid. The mesethmoid is a long bone composed of two slightly divergent 
posteriorly directed limbs. The frontals are inserted between these projections, 



e pal 




Fig. 27. Rhynchodercetis hakelensis (Pictet and Humbert). Neurocranium in dorsal 

view. 



ESPECIALLY MYCTOPHOIDS 61 

attaching to their medial edges. (A nasal component was not observed in the 
specimen and it may be that it is co-ossified with the mesethmoid.) The mesethmoid 
continues anteriorly in advance of the f rentals and ends in an acute point. The 
supraorbital sensory canal ran within the frontal. In the snout region the sensory 
canal appears to have been directed towards the lateral region of the mesethmoid 
(possibly supporting the assumption that a nasal component is present in the 
mesethmoid). The infraorbital sensory canal crossed the dorsal surface of the 
sphenotic beneath the lateral flange of the frontal and joined with the otic branch of 
the infraorbital sensory canal. This latter branch passed on to the dorsal surface of 
the sphenotic. 

The pterotic is an elongate bone forming the postero-lateral border of the skull- 
roof, and bearing a shallow longitudinal crest inclined medially anteriorly. The otic 
branch of the infraorbital sensory canal passed within this crest and anteriorly the 
tube for the canal is clearly visible. Posteriorly the pterotic projects backwards for 
a considerable distance beyond the occipital border. Ventro-laterally the pterotic 
forms most of the hyomandibular facet. The pterotic contacts the sphenotic 
anteriorly, and the sphenotic contributes towards the anterior region of the hyo- 
mandibular facet. Postero-medially the pterotic joins the epiotic and parietal in the 
open groove which may represent the post-temporal fossa. The epiotic is poorly 
shown but does exhibit a slight thickening medial to the groove which provided an 
articulatory area for the dorsal limb of the post-temporal. 

The parasphenoid is visible below the orbit and is considerably expanded into 
ventro-lateral flanges which presumably were associated with the medial edges 
of the endopterygoids. The lateral ethmoid delimits the anterior orbital region and 
is a large bone attached ventrally to the postero-dorsal region of the palatine. 

Jaws and suspensorium. The jaw suspensorium is inclined slightly forwards 
ventrally so that the quadrate condyle is positioned below the centre of the post- 
orbital part of the skull. The head of the hyomandibular is considerably elongated, 
and its rear edge bears a shallow crest against which the anterior edge of the pre- 
operculum rested. Approximately one-third of the way down the bone this crest 
is joined by a second crest from the anterior region of the hyomandibular head. 
The combined crest so produced shallows and disappears at the ventral extremity of 
the hyomandibular. The anterior extent of the hyomandibular is a thin sheet of 
bone which is covered by the metapterygoid. The quadrate is triangular and stout 
with the condyle directed antero-ventrally. Anteriorly the quadrate is attached to 
the lateral face of the ectopterygoid. The association of the quadrate with the 
metapterygoid is difficult to interpret, the latter appearing to overlap part of the 
hyomandibular and to bear an obliquely inclined crest anteriorly. The ectoptery- 
goid expands slightly below the orbit where it is more thickened and prominent. It 
is assumed that this region ventrally bears teeth although none were clearly observed. 
Further forwards, below the lateral ethmoid, the ectopterygoid contacts the postero- 
ventral part of the palatine. An endopterygoid is attached to the medial region of 
the ectopterygoid and is inclined medially to associate with the parasphenoid in the 
roof of the mouth. The palatine is excessively elongated, deepest posteriorly and 



62 UPPER CRETACEOUS TELEOSTS 

tapering in height anteriorly. Postero-dorsally it is supported by the ventral region 
of the lateral ethmoid, and more anteriorly it is associated medially with the lateral 
parts of both the mesethmoid and vomer. Anteriorly the palatines continue in 
advance of the mesethmoid and meet in the mid-line, terminating anteriorly in an 
acute point. Traces of teeth are seen along the length of the palatine. 

The dermal upper jaw is composed of two bones, the premaxilla and the maxilla. 
Each bone appears to contribute about half of the oral margin of the upper jaw. 
The premaxillae are produced for a considerable length in advance of the end of the 
palatines, and also meet in the mid-line. The premaxillae lie lateral to the palatines 
and have a smooth external surface. No traces of teeth are visible on the pre- 
maxillae. Only the posterior region of the maxilla is visible, i.e. where it enters the 
gape, and the anterior extent cannot be ascertained. The maxilla is a relatively 
shallow lamina of bone bearing a row of minute teeth. 

The mandible is long and very shallow, and only extended as far as the anterior 
end of the palatines. The articular facet is shallow and ill-defined and there is no 
distinct coronoid process, the rear edge of the articular merely sloping forwards to 
contact the dentary. The dentary is deeply indented posteriorly and this indenta- 
tion is occluded by the forward extent of the articular. A small angular element 
occurs below the articular facet. A lateral flange, running in a longitudinal direction, 
is present on the articular bone; when the jaws are closed the posterior end of the 
maxilla rests on the dorsal surface of this flange. Orally the dentary bears traces of 
teeth. No ornamentation occurs on the lower jaw. 

Opercular bones. The preoperculum is inclined forwards ventrally where it is 
associated with the rear edge of both the hyomandibular and the quadrate. It is a 
narrow bone with a slight ventral expansion. The operculum is very large and is 
as deep as it is long with no ornamentation. The suboperculum is as long as the 
operculum but is considerably shallower and its dorsal region lies medial to the 
ventral region of the operculum. An interoperculum is not present. 

Fins and fin supports. The fins are shown in the reconstruction, Text-figure 28. 
The pectoral girdle is hardly visible in the specimen, only traces of the cleithrum can 
be seen behind the opercular apparatus. The cleithrum is narrow with a consider- 
able vertical and horizontal extent, following the contour of the rear edge of the 
opercular bones. The pectoral fin insertion is high up on the flank, above the mid- 
line of the body. The fin consists of at least 10 branched and segmented rays. 

The pelvic bones are narrow fragile slips of bone positioned below the middle 
of the dorsal fin. The pelvic fin rays are feeble but all are branched and segmented, 
totalling at least 7 in number. 

The dorsal fin is short but relatively deep, the first two or three rays being longer 
than the basal length of the fin. The fin consists of about 6 rays which are relatively 
stout, though flattened, and only appear to be segmented at their extreme distal ends. 
The radial elements supporting the fin rays are short with an expanded articulatory 
region. The first radial also bears a slight anterior expansion. 

The anal fin is composed of at least 8 rays and is remote in position. 



ESPECIALLY MYCTOPHOIDS 63 

Vertebral column. The vertebral column is shown in the reconstruction, Text- 
figure 28. It consists of approximately 60 vertebrae, of which 30 appear to be 
caudal. The first 18 precaudal vertebrae are of uniform size, as long as deep. 
Between the 18th and the 22nd a slight increase in vertebral length occurs but after 
the 22nd the centra remain constant with a length slightly in excess of their depth. 
The length of each vertebra decreases in the region of the 20th caudal vertebra. 
Each centrum is mesially constricted and on the longer centra the constriction 
is most noticeable. The characteristic feature of the precaudal region is the 
presence on each centrum of two pairs of prominent transverse processes. The 
processes are directed ventro-laterally and occur at either end of the centrum. The 
anterior transverse process is larger than the posterior. The processes remain 
relatively constant in size on the first 15 precaudal vertebrae, but posterior to this 
they begin to elongate and reach their greatest length at about the 22nd precaudal 
vertebra, i.e. below the origin of the dorsal fin. Posterior to the 22nd vertebra the 
transverse processes gradually decrease in length and incline more ventrally. In the 
caudal region the anterior pair of transverse processes fuses ventrally to produce the 
haemal arches whilst the posterior pair appears to become zygapophyses. Ribs if 
present are very feeble and short, one or two centra do appear to have small fine ribs 
articulating with the underside of the larger, anterior, transverse processes. Long 
and low neural arches are present along the entire length of the column. Each 
neural arch occupies the whole dorsal extent of the corresponding centrum and gives 
rise to a feeble, posteriorly directed neural spine. These spines arise from the 
posterior region of each neural arch, and a prominent prezygapophysis is visible on 
the anterior region. The entire length of the body seems to have been provided 
with intermuscular bones, both epineurals and epipleurals, and possibly a certain 
number of epicentrals. Along a considerable part of the precaudal region in the 
mid-dorsal line there are several ossifications which are more substantial than the 
intermuscular bones. These may represent several median intermusculars com- 
pounded together, or, more probably, an ossified dorsal ligament. Just behind the 
occipital region of the head the anteriormost ossification appears to expand ventrally 
and connects with the neural spine of the first vertebra. 

Scales. The body is devoid of scales except for a single row on each flank, marking 
the course of the lateral line. All of the scales are based on a triradiate plan with a 
long forwardly projecting arm and two shorter divergent backwardly projecting 
arms. The interspaces between the smaller arms and the larger arm are variously 
filled with bone to give a more rounded outline which becomes more noticeable 
caudally. At the point of junction of the three arms a small backwardly projecting 
spine occurs : this is present on all the scales but is more pronounced and uncinate 
in the caudal region. 

Remarks. Arambourg's (1954) description of the type species Rhynchodercetis 
yovanovitchi (Text-fig. 29) differs in a few details from the above description of 
Rhynchodercetis hakelensis. Arambourg's supraorbital above the posterior end of 
the orbit does not appear to be a separate bone but merely the lateral extension of 
the frontal above the orbit. Here, as in specimens of other genera examined 



6 4 



UPPER CRETACEOUS TELEOSTS 












& 



1 

P 

X 



1 s 

aj en 






be 



B-s 



Si a" 

o 

s - 

S V 

««! 

8 

00 



S 00 

o 
S 
S 

o 



-« 
















ESPECIALLY MYCTOPHOIDS 65 

(notably Ichthyotringa delicata, Text-fig. 5, p. 14), the tube within the frontal which 
contained the supraorbital sensory canal passes medially from above the sphenotic 
and then curves anteriorly to pass along the dorsal surface of the frontal above the 
orbit. This tube often appears to demarcate the medial frontal region from the 
lateral frontal region, the lateral region thus having the appearance of a separate 
bone. Arambourg (1954 : 108) also described and figured separate nasals. I was 
unable to find these as discrete elements in Rhynchodercetis hakelensis, where they 
appear to have become incorporated with the mesethmoid or lost altogether, both 
situations occurring in other teleosts. The mandible does not always terminate at 
the posterior end of the premaxillae as Arambourg indicated for Rhynchodercetis 
yovanovitchi. In certain of his specimens it definitely extends further anteriorly, 
almost to the end of the rostrum in fact, thus agreeing with Rhynchodercetis hakelensis 
in which the mandible extends to the anterior end of the palatine. Finally Aram- 
bourg (1954 : 111) describes and figures a prominent interoperculum. I was unable 
to find this bone in any of his specimens and it does not appear to be present in 
Rhynchodercetis hakelensis. 

Rhynchodercetis gortanii (d'Erasmo) 
(Text-figs. 30A-C) 

1946 Leptotrachelus gortanii d'Erasmo : 76, pi. I, fig. 4. 

1946 Leptotrachelus sp. (cfr. L. hakelensis) d'Erasmo : 82. 

1952 Rhynchodercetis acutissimus d'Erasmo : 108, text-figs. 7-12, pi. 2, figs. 4-6, pi. 3, figs. 1-4. 

Diagnosis (emended). Rhynchodercetis of estimated standard length up to 22 cm. 
Vertebral column with 78-80 vertebrae of which 30 are caudal. Head equal to 
between a quarter and a third of the standard length. Maximum height of the head 
equal to one-sixth of the total head length. Dorsal fin with 8 rays, anal fin with 
16-20 rays. Pectoral fin with 10 rays, pelvic midway between the pectoral and the 
anal and composed of 8 rays. 

Holotype. The holotype of Leptotrachelus gortanii d'Erasmo, once in Bologna, 
could not be traced. The holotype of Rhynchodercetis acutissimus d'Erasmo is in the 
Museo Civico di Storia Naturale, Trieste. Both specimens come from the Lower 
Cenomanian of Comen, near Trieste. 

Material. Two trunk fragments from Comen in the Istituto di Geologia e 
Paleontologia della Universita, Bologna ; specimen number P.10913 in the 
B.M.N.H. ; and an incomplete specimen from the Natural History section, Prirodo- 
slovni Muzej Slovenije, Ljubljana, Jugoslavia. 

Remarks. Leptotrachelus gortanii was erected by d'Erasmo (1946) on a practically 
complete specimen from Gabrovizza in Comen, near Trieste. Together with the 
holotype d'Erasmo also used two trunk fragments from Dol Piccolo, also in Comen. 
The author has examined these latter specimens in Bologna but the holotype is 
missing. The holotype of Rhynchodercetis acutissimus d'Erasmo (1952) together with 
several other body fragments, all from Comen, are in Trieste. Unfortunately these 
specimens were not made available to me but subsequently photographs of this 



66 



UPPER CRETACEOUS TELEOSTS 



material have been studied. The original descriptions of d'Erasmo (1946 ; 1952) 
together with the body fragments in Bologna have convinced me that only one species 
is represented. A small head region, also from Comen, in Bologna, was noticed by 
d'Erasmo (1946 : 82). He ascribed this specimen to the genus Leptotrachelus, and 
indicated that it had affinities with the species Leptotrachelus hakelensis. It almost 
certainly is a specimen of Rhynchodercetis gortanii. 

Due to the lack of material, only a few features are described. All the features 
are well shown by the two specimens from Comen, one in the B.M.N.H. number 
P.10913, and the other in Ljubljana. Both specimens are of the mid-trunk region 
and exhibit both precaudal and caudal vertebrae. 

Description. Vertebral column. Selected vertebrae are shown in Text-figures 
30 A-C. Approximately 59 vertebrae are exhibited in the Ljubljana specimen, 
of which 39 are precaudal. (The type specimen of Rhynchodercetis acutissimus has 
80 vertebrae in all, of which 30 are caudal.) Anterior to the dorsal fin the specimen 
is flattened dorso-ventrally due to the prominent transverse processes having 
prevented twisting of the body during fossilization. However further back the 
dorsal fin and the pelvic fins have overriden this effect and the remainder of the body 
is preserved normally by lateral compression. The body was obviously long and 
very thin, and the precaudal region is characterized by the presence of two pairs of 




postz 



Fig. 30. Rhynchodercetis gortanii (d'Erasmo). Selected vertebrae, (a) Ventral view of 
precaudal vertebrae 26 and 27. (b) Lateral view of precaudal vertebrae 42 and 43. 
(c) Lateral view of precaudal vertebra, 50 and caudal vertebrae 1 and 2. (d) Lateral 
view of caudal vertebra 15. 

enormous transverse processes per centrum. The transverse processes are long and 
acutely pointed, but are flattened and extended where they contact the centrum. 
The anterior pair of transverse processes is almost twice as long as the posterior pair. 
The length of the anterior processes increases posteriorly and they attain their 
greatest length at about the 26th precaudal vertebra, (i.e. assuming that 80 verte- 
brae are present of which 50 are precaudal). The greatest length of the anterior 
transverse processes is in excess of the length of the centrum supporting them. The 
posterior pair of transverse processes are set at 90 ° to the long axis of the centrum 



ESPECIALLY MYCTOPHOIDS 67 

whereas the anterior processes incline slightly forwards. Pleural ribs articulate 
with the apices of the larger pair of transverse processes. Some of the smaller, 
posterior transverse processes appear to have thin bones articulating with them, but 
these are more likely to be some of the numerous intermuscular bones. Posterior to 
the 26th precaudal vertebra both the pairs of transverse processes begin to decrease 
in length so that by the end of the precaudal region, at about the 50th vertebra, the 
processes are much reduced and project ventrally as opposed to their more normal 
ventro-lateral inclination. The anterior pair of transverse processes fuses in the mid- 
ventral line to produce the haemal arches ; this seems to have occurred on the 51st 
vertebra. From the postero-ventral part of this haemal arch a protuberance occurs 
which on the succeeding caudal vertebrae becomes elongated, forming curved haemal 
spines which attain their maximum length on the 4th caudal vertebra. As has been 
pointed out, the haemal spine arises from the posterior region of the haemal arch, 
while the anterior region of the haemal arch projects forwards very slightly on either 
side to form ventral prezygapophyses. The posterior transverse processes also 
incline ventrally in the caudal region but do not fuse together, merely forming 
ventral postzygapophyses. Throughout the caudal region the ventral pre- and post- 
zygapophyses are extremely prominent. The entire length of the dorsal surface of 
each centrum is occupied by an elongated neural arch which is a separate ossification 
to the centrum, the suture between the two structures being clearly visible. The 
neural arches are long and low and bear neural spines as projections from the extreme 
postero-dorsal region. The anterior region of each neural arch is produced on 
either side to give prezygapophyses. The corresponding postzygapophyses are found 
on the postero-dorsal part of the centrum, posterior to the base of the neural spine. 
The neural spines themselves are not particularly prominent, being short, straight 
and relatively narrow. 



Genus PELARGORHYNCHUS von der Marck, 1858 

Diagnosis (emended). Dercetidae in which both the skull and body are 
elongated. Premaxillae do not extend in front of the mesethmoid. Mandible 
extends to the anterior region of the snout. Maxilla enters the gape and is toothed. 
Paired fins reduced in size, pectorals being low on the flanks, pelvics arising below the 
anterior end of the dorsal fin. Dorsal fin occupies the posterior half of the back, 
anal fin short and remote. A single row of heart-shaped flank scutes along the 
lateral line, and a dense intermediate squamation of small rhomboidal scales covers 
the remainder of the body. 

Type species. Pelargorhynchus dercetiformis von der Marck. 

Remarks. The genus was erected by von der Marck (1858) for the reception of 
several poorly preserved specimens from Sendenhorst, Westphalia. The genus is 
not recorded from any other localities. 



68 UPPER CRETACEOUS TELEOSTS 

Pelargorhynchus dercetiformis von der Marck 

(Text-fig. 31) 

1858 Pelargorhynchus dercetiformis von der Marck : 243. 

1954 Pelargorhynchus dercetiformis von der Marck ; Siegfried : 16, pi. 6, figs. 2-4. 
Siegfried lists earlier references. 

Diagnosis (emended). Pelargorhynchus of standard length not exceeding 49 cm. 
Vertebral column with 70 vertebrae. Head equal to one-fifth of the standard length. 
Maximum height of the head equalling one-fifth of the total head length. Dorsal fin 
with 64 rays, anal fin with 13 rays. Lateral line scutes elongate and heart-shaped. 

Holotype. The holotype appears to have been lost, but two additional specimens 
used by von der Marck (1858, 1863) are present in the Palaontologisches Institut der 
Westfalische Wilhelms-Universitat, Miinster, numbers 8488 and A1/3, both from the 
Upper Senonian of Sendenhorst. 

Material. The two specimens in Miinster together with a third specimen of a 
distorted trunk region in Miinster. These three specimens are all that could be 
traced of the species. The following description is very incomplete due to the poor 
preservation of the material. 

Description. Neurocranhim. The neurocranium is shown in dorsal view in 
Text-figure 31. The frontals extend back almost to the occipital border and are long, 
narrow bones which meet in the mid-line. The widest point of the skull-roof is 
between the sphenotics on a level with the hind end of the orbit. The frontals are 
densely ornamented with rows of tubercles raised on bony ridges. These ridges 
radiate in all directions from the centre of ossification above the rear end of the 
orbit. The sphenotic is visible on one side of the roof as a triangular region project- 
ing laterally from beneath the frontal. The frontals are greatly extended anteriorly 
and taper from in front of the orbits. The mesethmoid component of the skull is 
prominent and acutely pointed anteriorly, but more posteriorly is composed of two 
divergent laminae of bone flanking the anterior region of the frontals. In lateral 
view the orbit appears small, the majority of the skull length being preorbital. The 
parietals are small, transversely orientated strips of bone near to the occipital border 
and bounded anteriorly by the frontals. The supraoccipital is a small bone which 
contacts the frontals anteriorly and separates the parietals in the mid-line. A 
supraoccipital crest was not observed but preservation was poor in this region. 
Both parietals bear a dense ornamentation of tubercles but the dorsal surface of the 
supraoccipital is smooth. Behind the parietals and forming the occipital border of 
the roof, two small unornamented expanses of bone are visible, the epiotics. 
Postero-lateral to the frontals the pterotics border the skull-roof. The pterotics 
are elongate and relatively narrow although they do expand slightly in width 
posteriorly. The pterotic projects backwards beyond the level of the occipital 
border. Posteriorly a slight indentation of the pterotic is visible in the posterior 
border of the neurocranium just lateral to the epiotic and the parietal, and would 
seem to indicate the presence of a post-temporal fossa. The indentation may mark 



ESPECIALLY MYCTOPHOIDS 



69 




spo 



Fig. 31. Pelargorhynckus dercetiformis von der Marck. Neurocranium in dorsal view. 



70 UPPER CRETACEOUS TELEOSTS 

the opening of a completely roofed post-temporal fossa but is more probably the 
posterior portion of an unroofed or partially roofed fossa. 

Jaws and suspensorium. The suspensorium is vertical with the quadrate condyle 
lying below the occipital border. Little of the remainder of the palate could be 
made out except that the palatine appears to support a double series of hollow 
teeth. The palatine is closely associated with the anterior end of the snout and 
was in part supported by the mesethmoid. 

The dermal upper jaw consists mainly of an elongated, shallow premaxilla extend- 
ing from the extreme tip of the snout almost to the rear of the orbit. The pre- 
maxilla appears to contact its partner anteriorly in the mid-line. The lateral surface 
of the premaxilla is ornamented with tuberculated ridges, and the oral margin bears a 
uniform row of minute teeth. The maxilla enters the gape behind the premaxilla as 
a thin splint of bone, bearing a single row of small conical teeth. 

The mandible is elongated and very shallow with the dentary forming the whole of 
the oral border. The lateral face of the mandible is sparsely ornamented and the 
oral edge is covered with many small, acutely pointed, recurved needle-like teeth. 
The articular is more densely ornamented on its lateral face than the dentary and the 
facet for the reception of the quadrate condyle is shallow and ill-defined. 

The preoperculum is upright, narrow and with a slight basal expansion. The 
operculum is large and has a rounded posterior edge, whilst the suboperculum is 
much smaller. There are numerous branchiostegal rays in the region of the sub- 
operculum, but it would appear that the interoperculum was absent. 

Vertebral column and fins. The body is of a characteristic shape, the neck region 
being constricted and of no greater depth than the maximum depth of the head. 
The body widens gradually and attains its maximum depth at the level of the pelvic 
fin. The vertebral column is composed of approximately 70 vertebrae, the individual 
centra are mesially constricted and in the anterior region of the body are longer than 
they are deep. The centra decrease both in height and length as they near the 
caudal extremity. Apart from this little can be said concerning the vertebral 
column due to the complete covering of scales which masks the underlying structures. 

The pectoral fins are positioned quite low down on the flank and appear to consist 
of some 12 rays. The pelvic fins are larger and positioned midway along the body 
beneath the 32nd vertebra, and have about 7 rays, the first of which is unbranched. 
The dorsal fin occupies practically half of the length of the body and is composed of 
64 rays. The anal fin is remote in position with 13 rays, opposite the hind end of the 
dorsal fin. The caudal fin is prominent but not deeply cleft and consists of 19 
principal rays and approximately 12 accessory rays. 

Scales. The whole body is covered with a complete coat of small imbricate scales 
which are rhomboidal in shape. Along the course of the lateral line the scales are 
much enlarged, elongate and heart-shaped. Each lateral line scute bears a small 
posteriorly directed spine and an ornamentation of radiating tubercles. There 
might possibly be a second row of enlarged scutes on the body but this is not definite, 
despite Woodward's (1901 : 188) statement. 



ESPECIALLY MYCTOPHOIDS 71 

Suborder ENCHODONTOIDEI 

Diagnosis. Body relatively shallow and somewhat laterally compressed. 
Parietals separated completely by the supraoccipital ; post-temporal fossa roofed 
or unroofed. Orbitosphenoid and basisphenoid absent. Supraorbital and ant- 
orbital absent. Premaxilla with a fenestrated pedicel ; maxilla untoothed or finely 
toothed ; supramaxillae absent. Palatine bone thick and tumid with a terminal 
palatine tooth ; teeth absent on the vomer. Large number of branchiostegal 
rays. Interoperculum absent. Pectoral fins inserted low down on the flank ; no 
mesocoracoid arch. Pelvic fins abdominal or sub-thoracic. Vertebral elements 
incompletely fused ; epineurals present. Vertebrae variable in number, more than 
half the total being caudal. Fins without spines ; caudal with 19 principal fin-rays, 
17 branched. Caudal skeleton with second ural vertebra present as a half-centrum. 
Scales present only along the lateral line and the mid-dorsal line anterior to the 
dorsal fin. 



Family ENCHODONTIDAE Woodward, 1901 

Diagnosis (emended). Head deepened, especially posteriorly ; body may be 
deepened in thoracic region. Post-temporal fossa unroofed. Lower jaw long and 
deep behind the constricted symphysis ; articular facet visible from the lateral 
aspect. Operculum convex posteriorly and deeper than it is broad ; preoperculum 
without prominent ventral spine. Pectoral fins larger than pelvic fins and extremely 
low on body. Pelvic fins abdominal. No posterior extension of the cleithrum. 
Lateral line scales do not overlap ; mid-dorsal scutes reduced and not overlapping. 



Genus ENCHODUS Agassiz, 1835 

Diagnosis (emended). Enchodontidae in which the body tends to become 
shortened and deepened. Vertebral number may be reduced to 37, caudal region 
remaining relatively constant at 25 vertebrae. Pectoral fins only slightly larger 
than pelvic fins. Anal fin short based, arising behind dorsal fin. Dorsal fin short 
based and in the middle of the back. Mid-dorsal scutes reduced and few in 
number. 

Type species. Enchodus lewesiensis (Mantell) 

Remarks. The genus Enchodus, first erected by Agassiz (1835), is known by 
many species. The majority have been erected on fragmentary material consisting 
mainly of isolated teeth or fragments of the jaw bones (see Woodward 1901 : 190- 
205). Only a few species are represented by complete material and two of these 
are described in the following pages. The neurocranium and jaws of Enchodus 
faujasi Agassiz have been described in detail by Goody (1968). 



72 UPPER CRETACEOUS TELEOSTS 

Enchodus lewesiensis (Mantell) 
(Text-figs. 32-36, 38A) 

1822 Esox lewesiensis Mantell : 237, pi. 25, fig. 13, pi. 33, figs. 2-4, pi. 41, figs. 1-2. 

1835 Enchodus halocyon Agassiz, Feuill. : 55. 

1837 Enchodus halocyon Agassiz, 5 ; 1 : 64, pi. 25c, figs. 1-16. 

1850 Enchodus halocyon Agassiz ; Dixon : 373, pi. 30, figs. 20 & 27, pi. 31, fig. 11. 

1875 Enchodus halocyon Agassiz ; Geinitz : 226, pi. 41, figs. 5-20 (paytim). 

1883 Enchodus lewesiensis (Mantell) Sauvage : 480, pi. 12, fig. 16. 

1883 Enchodus halocyon Agassiz ; Gosselet : pi. 22, fig. 1. 

1888a Enchodus levesiensis (Mantell) ; Woodward : 315, pi. 1, fig. 5. 

1901 Enchodus lewesiensis (Mantell) ; Woodward : 191, pi. n, fig. 1. 

1901 Enchodus annectens Woodward : 195, pi. 11, figs. 4 & 5. 

1902 Enchodus lewesiensis (Mantell) ; Leriche : 139, pi. 4, fig. 13. 

1903 Enchodus annectens Woodward ; Woodward : 57. 

1903 Enchodus lewesiensis (Mantell) ; Woodward : 57, pi. 14, figs. 1-8, text-fig. 13. 

1906 Enchodus lewesiensis (Mantell) ; Leriche : 95, text-fig. 15. 

1908 Enchodus lewesiensis (Mantell) ; Priem : 52, pi. 1, fig. 10, pi. 2, fig. 2. 

1930 Enchodus annectens (Woodward) ; Maury : 99, fig. 4. 

1932 Enchodus lewesiensis (Mantell) ; Woodward : 160, fig. 246. 

1933 Enchodus lewesiensis (Mantell) ; Sarra : 30. 

1937 Enchodus lewesiensis (Mantell) ; Rode : 127, pi. 9. 

1964 Enchodus lewesiensis (Mantell) ; Danil'chenko : 411, fig. 127. 

Diagnosis (emended). Enchodus in which the length of the head exceeds its 
depth at the occiput ; maximum depth of the body behind the occiput estimated as 
equal to the length of the head. Dermal bones of the head ornamented with radiat- 
ing tuberculated ridges. Mandibular teeth six in number of which the anteriormost 
is the longest ; both hyopalatine and mandibular teeth long, thin and needle-like. 
Palatine tooth fractionally longer than the first tooth on the ectopterygoid. 
Mandibular symphysis much constricted ; mandible deepens posteriorly so that the 
maximum depth is equal to one-third of its total length. Maxilla long and thin and 
probably untoothed. 

Syntypes. Specimens in the B.M.N.H., numbers 4004, 4049, 4157, 4160, 4180, 
4181, 4183, 4184, 4198, being a collection of teeth and jaw fragments from the 
Turonian and Senonian of south-east England. 

Material. Specimens in the B.M.N.H. totalling some 20 fragmentary pieces, 
listed by Woodward (1901 : 192). 

Remarks. The type species is only known by imperfect specimens of the head. 
The series of type specimens are those figured by Mantell (1822) and Agassiz (1837), 
all from the Chalk of Sussex. The following description is based mainly on 
specimen number 4001 in the B.M.N.H. from the Middle Chalk of Lewes in Sussex, 
described by Woodward (1901 : 192) as a very imperfect small skull. The specimen 
was prepared in acetic acid and a small practically complete neurocranium was 
obtained. The description of the jaws was compiled from several fragmentary 
specimens, notably specimen number P. 5415 which had already been partially 
described and figured by Woodward (1888a : 315, pi. 1, fig. 5). 



ESPECIALLY MYCTOPHOIDS 



73 



Description. Neurocranium. Dorsal, ventral, lateral and posterior views of the 
neurocranium are shown in Text-figures 32-35. The cranial roof is flattened with its 
widest point behind the orbits in the sphenotic region. The cranium is relatively 
deep in the posterior region but shallower anteriorly. The orbit is enormous, 
occupying more than half the length of the neurocranium, whilst the cranial cavity 
itself is very small. 




5mm 



st.s.r. 



Fig. 32. Enchodus lewesiensis (Mantell). Neurocranium in dorsal view. The broken 
line on the right side of the figure indicates the course of the sensory canals. 



74 UPPER CRETACEOUS TELEOSTS 

The frontals form almost the whole of the skull-roof, meeting in the mid-line in a 
distinct suture. A deviation of the suture to the right side of the roof is very 
noticeable in the posterior third of the frontal region. Posteriorly the frontals end 
near to the occipital border overlapping the anterior regions of the remainder of the 
roofing bones. The medial region of the roof is flat and relatively unornamented, 
and between the orbits there is a slight inter-orbital depression. The more lateral 
region of each frontal bears extremely prominent ridges which run in longitudinal 
and lateral directions from the centre of ossification. The major ridges bear an 
ornamentation of small bony tubercles. Several of the most prominent ridges 
pass forwards towards the snout region ; others pass laterally to terminate on the 
upper margin of the orbit. The ridges which extend posteriorly are arranged into 
a fan, the most lateral ridge passes straight back along the upper margin of the 
internal wall of the post-temporal fossa while the most medial ridge passes in 
towards the mid-line. The postero-lateral region of each frontal displays a V-shaped 
indentation of the upper ornamented layer. This indentation marks the forward 
extension of the unroofed post-temporal fossa. The frontal itself extends into the 
fossa forming the anterior regions of both the medial and lateral walls as well as 
the floor. The posterior half of the fossa is composed medially of parietal and 
laterally of pterotic. The frontal is applied to the dorsal surface of the sphenotic, 
this latter bone producing the upper hind limit of the orbit. Anteriorly the 
frontals taper from the orbits and contact the mesethmoid. 

The supraoccipital appears on the surface of the neurocranium as a small median 
bone with its anterior region overlain by the backward extension of the frontals, thus 
separating the parietals. Its internal extent is probably far greater than the out- 
ward appearance would indicate. As well as contacting the frontals and parietals, 
it also contacts the epiotics postero-laterally and the exoccipitals on the posterior 
face of the neurocranium. The supraoccipital crest is small, arising from the 
posterior face and not extending above the level of the skull-roof. On the surface of 
the supraoccipital, anterior to the origin of the crest, there is a shallow transverse 
groove crossing the complete width of the bone. 

The parietals are composed of two distinct regions, an anterior roughly triangular 
portion which is heavily ornamented and closely associated with the frontals ; and 
a smaller postero-lateral portion which is smooth and unornamented. In this 
posterior, smooth region, the epiotic and parietal join in the floor of a distinct 
transverse groove. The grooves are continuous with the transverse groove on the 
supraoccipital. Laterally the parietal forms the posterior half of the medial wall 
of the post-temporal fossa. 

The pterotic appears on the skull-roof lateral to the posterior region of the frontal. 
Anteriorly the pterotic is covered by a lateral extension of the frontal in the region 
of the post-temporal fossa. The pterotic is produced into a prominent vertical crest 
which forms the lateral wall and floor of the post-temporal fossa, and also part of the 
medial wall of the dilatator fossa. The pterotic contacts the sphenotic anteriorly 
and the suture is visible within the dilatator fossa. On the postero-dorsal edge of 
the pterotic crest a large foramen leads into a tube within the pterotic crest. A 



ESPECIALLY MYCTOPHOIDS 75 

distinct transversely orientated oval fenestra is present at the hind end of the post- 
temporal fossa. This enters the neurocranium beneath the post-temporal fossa, and 
is bordered medially by parietal and epiotic and laterally by pterotic. The possible 
significance of this oval fenestra entering the cranial cavity is considered by Goody 
(1968 : 225). The pterotic crest bears a notch and a groove postero-laterally just in 
front of the hind end of the crest, the groove orientated postero-ventrally. The 
dilatator fossa, which housed the dilatator muscle of the operculum, is a vertically 
arranged facet produced by excavations of the sphenotic anteriorly and the pterotic 
posteriorly. The dilatator fossa is lateral in position and below it is the hyoman- 
dibular facet. This latter is an elongated facet consisting of a straight groove 
arising anteriorly on the sphenotic and being continued posteriorly on the pterotic. 
The pterotic therefore contacts the sphenotic anteriorly within the dilatator fossa 
and the hyomandibular facet. 

The sphenotic enters into the composition of the skull-roof and also forms the 
postero-dorsal corner of the orbit. Dorsally it projects laterally from beneath the 
frontal, and the surface is unornamented. The posterior surface of the lateral 
projection is excavated somewhat to produce the anterior end of the dilatator fossa. 
From the outermost region of the lateral extension a stout strut passes ventro- 
medially to meet a corresponding upward strut from the prootic. The anterior end 
of the hyomandibular facet is located behind this strut and below the dilatator fossa. 
In the orbit the sphenotic produces the major part of the hind wall, contacting the 
prootic ventrally, the pleurosphenoid medially and the frontal dorsally. The main 
infraorbital sensory canal crossed the dorsal surface of the sphenotic on which it 
appears to have connected with the otic branch of the infraorbital canal from the 
pterotic. 

The epiotic forms the postero-dorsal part of the cranial roof medial to the post- 
temporal fossa. The epiotics are separated in the mid-line by the supraoccipital and 
they contact the parietals anteriorly within the transverse groove already noticed. 
Postero-dorsally the epiotic is in the form of a raised rounded knob of bone which 
provided the articulatory surface for the dorsal limb of the post-temporal. On the 
posterior face of the skull the epiotic meets the supraoccipital and exoccipital 
medially and the intercalar ventrally. Within the extreme posterior end of the 
post-temporal fossa the epiotic contacts the parietal and pterotic where these three 
bones form the borders of the small oval fenestra entering the cranium below the 
post-temporal fossa. 

The supraorbital sensory canal ran in a tube of bone within the frontal, opening 
anteriorly on the edge of the frontal behind the mesethmoid region. Two sub- 
sidiary branches of the main supraorbital sensory canal arose from the main canal in 
the region of the centre of ossification. One branch passed antero-medially and 
opened on to the interorbital region of the roof ; the second passed postero-medially 
and opened on to the posterior surface of the frontal, from where it appears to have 
continued on the surface towards the parietal. The infraorbital sensory canal 
passed across the dorsal surface of the sphenotic within the groove present on that 
bone. It entered the lateral region of the frontal just posterior to the centre of 



76 



IH'PER CRETACEOUS TELEOSTS 



ossification. The otic branch of the infraorbital sensory canal probably connected 
with the main infraorbital canal within the frontal, or possibly on the dorsal surface 
of the sphenotic. It entered the foramen at the posterior end of the pterotic crest 
and passed anteriorly within the crest. The notch and subsequent groove seen on 
the postero-lateral region of the pterotic crest probably indicate the point at which 




a.t.f.c 



bo x 



5mm 



f.vpt 



Fig. 33. Enchodus lewesiensis (Mantell). Neurocranium in ventral view. 



ESPECIALLY MYCTOPHOIDS 



77 



the preopercular sensory canal entered the pterotic to contact the otic branch of the 
infraorbital canal. The presence of a complete supratemporal commissure is 
indicated by the continuous transverse groove on the parietals and epiotics and the 
supraoccipital. This groove ends laterally above the posterior end of the post- 
temporal fossa. Possibly the sensory canal passed laterally in a superficial position 
and joined with the otic branch of the infraorbital canal immediately posterior to 
the pterotic crest. 

The mesethmoid is prominent and well ossified, bearing a pair of lateral wings 
which are fused to the dorsal surface of the vomer. The composite wings thus 
formed produce an elongate articulatory region which fits into a correspondingly 
elongated facet on the internal face of the anterior palatine region. The anterior 
edge of the mesethmoid is straight and transversely arranged. The pedicels of the 
premaxillae rested upon this edge. 

The lateral ethmoids are well ossified both posteriorly and ventrally. The ventral 
region is pitted and excavated and was in life probably capped by cartilage. This 
region of the lateral ethmoid supported the posterior end of the palatine. Dorsally 
the lateral ethmoids are splayed apart and attach to the underside of the frontals. 

The vomer is closely adherent to the ventral surface of the mesethmoid, although 
its actual outline is difficult to determine. Teeth are definitely not present on the 
vomer. 

The parasphenoid is continuous for practically the entire length of the neuro- 
cranium. The ascending processes of the parasphenoid are inclined at 90° to the rest 
of the bone. Throughout the orbital region the parasphenoid is of uniform shape 




Fig. 34. Enchodus lewesiensis (Mantell). Neurocranium in lateral view. 



7 8 UPPER CRETACEOUS TELEOSTS 

with very slight expansions on either side, in the form of ventro-lateral flanges. The 
ventral surface has a shallow longitudinal groove, whereas the dorsal surface has 
a very low longitudinal ridge. Each ascending process of the parasphenoid meets 
the prootic laterally, and its anterior edge, bordering the myodome, is thickened. 
Posteriorly the ascending process is considerably thinner where it forms the wall 
of the myodome. In the antero-ventral region of the ascending process is a fora- 
men through which the internal carotid artery passed. The posterior part of the 
parasphenoid articulates with the ventro-lateral edges of the basioccipital, and ends 
posteriorly immediately below the occipital condyle. The postero-ventral region of 
the parasphenoid takes the form of a median keel. This keel occludes a certain 
amount of the lumen of the posterior myodome and the myodome does not open 
posteriorly. 

The prootic merits particular attention due to its complexity. The ventral part 
of the bone consists of two vertical laminae joined anteriorly. The inner sheet 
contacts the basioccipital behind and separates the myodome from the otolith 
chamber. The outer sheet forms the wall of the anterior part of the otolith chamber 
and meets the parasphenoid below and the basioccipital and exoccipital behind. 
The inner sheet curves medially and fuses with its partner from the opposite side 
forming the prootic bridge which separates the myodome from the cranial cavity. 
Posterior to the prootic bridge the otolith chambers communicate with the cranial 
cavity through a large oval fenestra. Above the prootic bridge the orbital face of 
the bone is roughly triangular, the apex of the triangle being directed ventrally. 
This region contacts the sphenotic dorso-laterally and the pleurosphenoid dorso- 
medially. Medially the edge of the prootic above the prootic bridge forms the 
margin of the lower part of the optic fenestra. The dorsal region of the lateral 
surface of the prootic is inclined dorso-laterally at about no to the ventral region. 
This upper region meets the exoccipital posteriorly, and the pterotic and sphenotic 
dorsally just below the hyomandibular facet. The prootic does not enter into the 
composition of the hyomandibular facet, which is produced solely from sphenotic and 
pterotic. 

The prootic contains the trigemino-facialis chamber (Text-fig. 38A), which is 
divided by bone into a medial pars ganglionaris and a lateral pars jugularis. The 
former is a recess on the medial face of the bone, within the cranial cavity, lateral to 
the anterior part of the prootic bridge, and which housed the ganglia of the trigeminal 
and facial nerves. The pars jugularis is an extremely short horizontal canal on 
the lateral face of the prootic lateral to the prootic bridge, and in the angle between 
the dorsal and ventral parts of the lateral surface of the bone. There is a single 
foramen leading from the pars ganglionaris into the pars jugularis, this being the 
facial foramen which transmitted the hyomandibular trunk of the facial nerve. 
The palatine branch of the facial nerve did not enter the pars jugularis but passed 
through a small foramen medial to the facial foramen into the myodome. The pars 
jugularis has two external openings, one posterior and one anterior. Through this 
very short canal passed the jugular vein and the orbital artery. The posterior 
opening also transmitted the hyomandibular branch of the facial nerve. The 



ESPECIALLY M YCTOPHOIDS 



79 



anterior opening of the canal is practically within the hind limit of the orbit. The 
remainder of the branches of the trigeminal and facial nerves did not enter the 
pars jugularis, but all passed through a single foramen in the hind wall of the orbit 
dorso-medial to the anterior opening of the pars jugularis. A notch occurs on the 
medial edge of the bone above the prootic bridge, this might indicate the course of 
the oculomotor nerve. 

The pleurosphenoids, in the postero-dorsal region of the orbit, are separated in the 
mid-line by the large optic fenestra. Each bone meets the frontal dorsally and the 
prootic and sphenotic posteriorly. Beneath the frontals the bone is splayed apart 
and the superficial sheets of bone produced fail to contact the frontals medially. 
This area was in life presumably filled by cartilage. 

The occipital condyle is circular and deeply concave. Ventrally it is formed of 
basioccipital. This single median bone passes forwards above the parasphenoid to 
meet the prootic anteriorly and the exoccipitals dorsally. In the angle between the 
condyle and the lateral region of the basioccipital there is a small group of pits and 
ridges. Internally the basioccipital forms the myodome roof and the floor to the 
otolith chamber. The basioccipital passes antero-dorsally to meet the prootics, 
separating the otolith chambers from the myodome. Dorsally the basioccipital 
bears a slight median crest which serves partially to separate the otolith chambers. 
The otolith chambers themselves extend anteriorly and end in a recess just posterior 
and ventral to the internal opening of the facial foramen. 



soc 




Fig. 35. Enchodus lewesiensis (Mantell). Neurocranium in posterior view. 



80 UPPER CRETACEOUS TELEOSTS 

The exoccipitals meet in the mid-line of the posterior face of the neurocranium 
both above and below the foramen magnum, completely enclosing it. Ventrally each 
exoccipital forms one dorso-lateral third of the occipital condyle. From the condylar 
region two struts of bone pass outwards, one laterally and one dorso-laterally ; the 
lateral strut meets the pterotic, but this is overlain by the superficial intercalar. 
The ventral edge of this strut, near to the condyle, has a large foramen within it. 
This foramen transmitted the vagus nerve and possibly also the glossopharyngeal 
nerve. The dorso-lateral strut meets the epiotic and is continued upwards by 
the latter bone. Lateral to the foramen magnum the exoccipital is pierced by one or 
two very small foramina which appear to have transmitted occipital nerves. On the 
lateral face of the skull the exoccipital forms the postero-dorsal wall of the otolith 
chamber and meets the basioccipital ventrally, the prootic anteriorly, and the 
pterotic dorsally. Dorsally the exoccipital contributes to a large concave depression 
in the cranial wall which served to house musculature associated with the branchial 
apparatus. Internally, in the floor of the foramen magnum, the exoccipitals meet 
in the mid-line to form a shelf. The anterior edge of this shelf forms the posterior 
margin of the fenestra between the otolith chambers and the cranial cavity. 

The intercalar is a moderately large bone incorporated into the cranial wall. It 
forms the main part of the lateral strut from the exoccipital and is wrapped around 
the postero-ventral corner of the braincase. It contacts the pterotic dorsally below 
the hind end of the hyomandibular facet. On the posterior face of the skull the 
intercalar bears a small projection which provided the articulatory region for the 
ventral limb of the post-temporal. 

Hyopalatine bones. A reconstruction of the hyopalatine bones is shown in medial 
view in Text-figure 36. The hyomandibular is relatively deep and robust and is 
vertical in position. The head of the hyomandibular is single and elongate and fits 
into the grooved hyomandibular facet. The opercular process is situated about 
mid-way down the bone. Laterally the hyomandibular bears a shallow crest which 
runs obliquely back from the anterior region of the head of the bone. The anterior 
extent of the bone is considerably thinner and was covered in part by the metaptery- 
goid. 

The quadrate is large, stout and triangular. The condyle is ventral in position 
and is arranged transversely. The posterior region of the upper border bears a deep 
indentation which marks the upper margin of the groove on the medial face of the 
bone which housed the symplectic. The ventral region of the preoperculum rested 
against the posterior edge of the quadrate, which is considerably thickened. The 
anterior quadrate region is indistinct but it appears to be sandwiched between 
posterior extensions of the ectopterygoid, or both the ectopterygoid and the endo- 
pterygoid. 

The ectopterygoid is bent through an angle of about 30 ° towards its posterior 
end where it overlaps the quadrate laterally. Both the ectopterygoid and endo- 
pterygoid are in the form of elongated laminae, which are fused along their ventral 
edges thus leaving a long dorsal V-shaped trough between them. The ectopterygoid 
is relatively shallow, the endopterygoid being much deeper. Anteriorly the ecto- 



ESPECIALLY MYCTOPHOIDS 




art 



Fig. 36. Enchodus lewesiensis (Mantell). Hyopalatine bones and mandible of the 

left side in medial view. 



pterygoid expands medially in front of the endopterygoid, so that both sides of 
the V-shaped trough are formed of ectopterygoid. The trough widens anteriorly 
into a U-shape, and the postero- ventral region of the palatine rests within it. 
On the ventro-lateral region of the palatine bone there is a longitudinal ridge pro- 
jecting laterally and ventrally. The ectopterygoid spreads anteriorly, ventral to 
this ridge and articulates with its undersurface. Ventrally the ectopterygoid 
bears a row of large teeth which decrease in size posteriorly. The anteriormost 
ectopterygoid tooth arises from the ventral edge of the bone. Farther back along 
the ectopterygoid, the endopterygoid projects below the suture line between the 
two bones so that the more posterior teeth of the ectopterygoid series tend to arise 
from the ventro-lateral surface. The bases of the teeth are angled in such a manner 
as to make the main body of each tooth hang vertically downwards. There are 
approximately 7 functional teeth present on the ectopterygoid at any one time, and 
replacement teeth are visible close to the functional ones. All of the teeth are 
elongated and needle-like with well defined, but not particularly expanded bases. 
The teeth are laterally compressed to give rise to two cutting edges which are not 
strictly anterior and posterior, but set somewhat obliquely to the long axis of the 
ectopterygoid, and have a slightly recurved aspect. 

The endopterygoid is visible from the lateral aspect as a small strip of bone below 
the region of fusion with the ectopterygoid. The endopterygoid bears a marginal 
row of minute teeth. On the internal face of the endopterygoid several more rows 
of minute teeth are seen radiating forwards and upwards. The more dorsal region of 
the endopterygoid inclines medially where it forms the roof of the mouth. 

The palatines are characteristically shaped and composed of dense bone. The 
characteristic feature is the possession of a single, enlarged, terminal tooth. This 



82 UPPER CRETACEOUS TELEOSTS 

tooth is the longest tooth in the upper jaw and forms the continuation, anteriorly, of 
the tooth row on the ectopterygoid. The base of the tooth merges imperceptibly 
into the main body of the bone and is arranged at 90 ° to it. The tooth is, in all 
respects, identical to the pterygoid teeth, being laterally compressed to produce 
cutting edges. From its extreme anterior position and the fact that there is no other 
tooth base or any indication of a second tooth, it is assumed that this tooth is not 
replaceable and remained fully functional throughout life. On the dorso-medial 
region of the anterior end of the palatine there is a facet which housed the lateral 
wing of the mesethmoid and vomer. Posteriorly the palatine fits into the dorsal 
groove of the ectopterygoid and dorsally is extended to produce a facet for the 
reception of the ventral region of the lateral ethmoid. The lateral face of the 
palatine has a shallow, oblique, ill-defined groove in which the most anterior region 
of the maxilla was contained. 

Dermal upper jaw. The dermal upper jaw consists of two bones, the maxilla and 
the premaxilla (Woodward, 1903, fig. 13). The premaxilla is particularly prominent 
with an anterior and a lateral region arranged at practically 90 ° to each other. 
There is an enlarged premaxillary pedicel anteriorly which contacts its opposite 
partner in the mid-line. The pedicels lie in the same plane as the anterior region of 
the skull-roof, and each pedicel contains a large oval foramen within it. The pre- 
maxillary pedicels articulate with the anterior edge of the mesethmoid. The lateral 
face of the bone is relatively smooth although grooves are present. One groove runs 
parallel with the pedicel at the anterior end of the lateral face, and a second very 
shallow groove runs along the oral border of the bone. The oral border itself bears a 
single row of marginal, acutely pointed teeth, approximately 20 in number. 

The maxilla is long and slender, the head being unexpanded and incurved. The 
head of the maxilla rests in the groove on the lateral and dorso-lateral surfaces of 
the palatine. The maxilla slopes posteriorly and ventrally and was presumably 
attached to the premaxilla ligamentously. The maxilla is straight, of uniform shape, 
entering the gape of the mouth at the posterior end of the premaxilla. The maxilla 
shows no signs of having been provided with teeth. 

Mandible. The mandible is shown in medial view in Text-figure 36. It is very 
long and deep, with the symphysial region much constricted and irregular in shape. 
The whole upper margin of the mandible is formed of dentary as well as the anterior 
two-thirds of the lower edge. The dentary is a large V-shaped bone with the apex 
toward the symphysis. The symphysis is deepened slightly by the presence of 2 or 
3 small, finger-like ventral processes. Ventrally the bone is much thickened. The 
oral edge of the dentary gives rise to an internal tooth bearing flange. Anteriorly 
this flange contacts the lower thickened region demarcating an internal V-shaped 
space. This space presumably housed the remains of Meckel's cartilage. Teeth are 
present in two rows, a marginal row of minute, acutely pointed teeth approximately 
30 in number ; and an inner row of enormous teeth. These latter teeth are very 
similar to those found on the palato-pterygoid. The row consists of about 8 func- 
tional teeth. The anterior tooth of the series arises just behind the symphysis and is 



ESPECIALLY MYCTOPHOIDS 83 

considerably longer than the rest of the teeth although of the same form. The 
remainder of the teeth decrease in size posteriorly. 

The articular forms the posterior mandibular region, and contains the concave, 
transversely orientated articular facet. There is a small retroarticular process 
behind the condyle which bears a deep groove passing around the condylar facet. 
This groove extends forwards ventro-laterally along the articular and dentary, and 
housed the mandibular sensory canal. The anterior and dorsal regions of the 
articular are in the form of a thin sheet of bone inserted on to the internal face of the 
dentary. 

Ornamentation on the mandible is sparse ; a few longitudinal ridges occur at the 
ventral angle of the jaw radiating forwards and upwards on the articular. From the 
symphysial end of the dentary several ridges run backwards and downwards. The 
most prominent ridges are ventro-lateral in position. 

Addendum. A small very imperfect neurocranium from the English Chalk was 
also prepared in acid, B.M.N.H. number P. 6461 (Text-figs. 37, 38B). This specimen 
had been noticed by Woodward (1901 : 193 and 1903 : 58) and included by him in 
the species Enchodus lewesiensis. This specimen would clearly seem to represent a 
species of Enchodus but there are several features which differ from those already 
noticed in the description of Enchodus lewesiensis. However since so little of the 
neurocranium is preserved and nothing of the remainder of the body, the erection of 
a new species is unnecessary. It will suffice here to point out the differences that 
are apparent. 

Comparing the dorsal neurocranial views (Text-figs. 32 and 37), the differences in 
the relative proportions of the roofing bones are obvious. The actual cranial cavity 
is more elongated in P. 6461 than in 4001 (the specimen used in the description of 
Enchodus lewesiensis). In fact the whole neurocranium would appear to have been 
longer and narrower in P.6461. The post-temporal fossae are similar in both speci- 
mens but the dilatator fossa of P. 6461 is larger and clearly visible dorsally (as it is in 
Enchodus faujasi ; Goody, 1968, fig. 1). In 4001 the dilatator fossa faces laterally 
and is hardly visible from the dorsal aspect. The epiotic of P.6461 is more pro- 
nounced and backwardly projecting, and lateral to it, at the rear end of the post- 
temporal fossa, the fenestra entering the neurocranium below the fossa is larger 
than in 4001. 

The most interesting difference between the two concerns the trigemino-facialis 
chamber in the prootic. This complex has been considered in some detail by 
Patterson (1964 : 434) where he deals with the changes that have occurred in the 
chamber during the evolution of the acanthopterygians. In P.6461 (Text-fig. 38B) 
the pars jugularis has three openings to the exterior. An anterior one in the hind 
face of the orbit, a posterior one opening on the lateral face of the prootic and a third 
one opening midway between the anterior and posterior openings. This last 
mentioned opening is large and the wall of the pars jugularis posterior to it is reduced 
to a narrow splint of bone. This opening would have transmitted the hyomandibular 
branch of the facial nerve dorsally, and presumably the orbital artery entered 
ventrally (the facial foramen leading into the pars ganglionaris in the medial wall of 



UPPER CRETACEOUS TELEOSTS 



so. s.c 




inf. s.c 



so.s.c 



o inf. s.c 



st.s.c 



5mm 



f.m 



Fig. 37. Enchodus sp. Neurocranium in dorsal view. Only the restored outline of the 
anterior region is shown. The broken line on the right side of the figure indicates the 
course of the sensory canals. From B.M.N.H. number P. 6461. 



ESPECIALLY MYCTOPHOIDS 



85 



the pars jugularis is visible through this gap in the lateral wall). Vestiges of the 
narrow splint of bone forming the jugular canal are visible in 4001 (Text -fig. 38 A) 
where they appear as small projections on the surface of the prootic both above and 
below the region of the jugular passage. The remainder of the trigemino-facialis 
chamber is similar in both specimens, with the trigeminal foramen opening into the 
hind wall of the orbit and not into the pars jugularis. In P. 6461 a further small 
elongated foramen is seen dorso-medial to the trigeminal foramen and may have 
transmitted either the profundus or the oculomotor nerve. 

If in fact specimen P. 6461 does belong to the genus Enchodus then the degeneration 
of the wall of the jugular canal and the confluence of the hyomandibular and jugular 
openings can occur within a single genus. This process has been shown by Patterson 
(1964) to be a general trend within the teleosts associated with an overall ad vane - 
ment in bodily structure. 



hm.VII 



a.t.f.c 




p.t.f.c 



Fig. 38. Prootic bones in antero-lateral view, (a) Enchodus lewesiensis (Mantell). 
From B.M.N.H. number 4001. (b) Enchodus sp. From B.M.N.H. number P.6461. 



Enchodus marchesettii (Kramberger) 

(Text-figs. 39-42) 

1895 Eurygnathus marchesettii Kramberger : 34, pi. 7, figs. 2, 2a. 

1901 Enchodus marchesettii (Kramberger) Woodward : 200. 

1903 Enchodus marchesettii (Kramberger) ; Hay : 421, pi. 30, figs. 2, 3. 

1952 Enchodus marchesettii (Kramberger) ; Arambourg : 232, fig. 50. 

Diagnosis (emended). Enchodus species of standard length not exceeding 18 cm. 
The length of the head only just greater than the depth at the occiput, and just less 
than the maximum depth of the body. Head with opercular apparatus equal to one- 
third of the standard length. Mandible long and deep, its maximum depth being 
less than one-third of its total length. Teeth much smaller than in the type species. 
Maxilla thin and toothless. Operculum almost twice as deep as broad ; sub- 
operculum very deep. Dorsal fin consisting of 16 rays ; anal fin with 20 rays and 



86 UPPER CRETACEOUS TELEOSTS 

arising behind the dorsal, nearer to the caudal than to the pelvics. Thirty-seven 
vertebrae of which 24 are caudal. 

Holotype. Once in the Museo Civico di Storia Naturale, Trieste, present loca- 
tion unknown. The holotype was from the Middle Cenomanian of Hakel, Lebanon. 

Material. Specimens in the B.M.N.H., numbers P.4748, P. 9250, P.47316, from 
the Middle Cenomanian of Hakel and all prepared by the transfer method in acetic 
acid. 

Remarks. The species is known solely by flattened specimens from Hakel and 
Hajula in the Lebanon. Woodward (1901 : 200) noted this species and stated that 
it was not clearly distinguishable from Enchodus longidens. Hay (1903 : 421) 
separated Enchodus marchesettii as a distinct species. The neurocranium was not 
clearly observable in the specimens examined, but it does not appear to have 
differed to any noticeable extent from that of Enchodus lewesiensis, or that of 
Enchodus faujasi (Goody, 1968). Arambourg (1952 : 232, fig. 50) shows a photo- 
graph of the caudal skeleton and compares it with that of Enchodus libycus Quaas 
in which he indicates two large plate-like hypurals. This comparison of the two tail 
skeletons is open to some doubt when compared with the description and figure of 
the caudal skeleton presented below. 

Description. Neurocranium. The frontals are visible merging anteriorly with 
the mesethmoid. Immediately posterior to this the lateral edge of the frontal bears 
a large elongated pore which is the anterior opening of the tube which in life con- 
tained the supraorbital sensory canal. The lateral region of the frontal is visible 
above the orbit and bears an ornamentation of tuberculated ridges. Above the 
sphenotic immediately posterior to the hind end of the orbit the postero-lateral 
region of the frontal is indented, indicating the anterior demarcation of the forward 
extent of the unroofed post-temporal fossa. This indentation is only in the upper 
ornamented layer of the frontal, since the frontal itself forms part of the floor of the 
anterior region of the fossa. The frontal overlaps part of the dorsal surface of the 
sphenotic. 

The cranium is relatively deep in the posterior region but becomes shallower 
anteriorly. The orbit is very large, occupying about half the length of the neuro- 
cranium, while the cranial cavity itself is rather small. The frontal meets the 
parietal posteriorly in the medial wall of the post-temporal fossa. Immediately 
behind the parietal there is a rounded knob of bone which may have been separated 
partially from the parietal by a groove, this is the epiotic forming the occipital 
border of the skull-roof. The crest on the pterotic can be seen crushed slightly into 
the post-temporal fossa. This crest forms the lateral wall of the fossa, as well as the 
medial demarcation of the dilatator fossa. The suture between the sphenotic and 
pterotic is visible within the fossa. At the posterior end of the pterotic crest there is 
some evidence of a pore leading into a tube within the crest which transmitted the 
otic branch of the infraorbital sensory canal. The upper boundary of the hyo- 
mandibular facet, formed of pterotic and sphenotic, is visible below the dilatator 
fossa. The sphenotic forms the upper posterior limit of the orbit and extends 



ESPECIALLY MYCTOPHOIDS 



87 



ventro-laterally. The parasphenoid is visible beneath the orbit as an almost 
straight rod of bone with no lateral expansions. From the underside of the frontals, 
above the orbit, the lateral ethmoid is produced downwards and connects with the 
hind end of the palatine. The frontals bear an ornamentation of prominent ridges, 
the major ridges running longitudinally, especially above the orbit, and following the 
course of the supraorbital sensory canal. The mid-region of the frontals is smooth 
and unornamented ; some small ridges appear above the sphenotic and pass 
laterally to the upper margin of the orbit. 

Infraorbital bones. The infraorbitals are shown in the lateral view of the skull, 
Text-figure 39. There are four narrow elongate infraorbitals and an elongated 
anterior lachrymal. The lachrymal is the largest bone in the series and is longer 
than deep. The antero-dorsal region of the lachrymal is produced dorsally, forming 
the anterior limit of the orbit. This dorsal projection articulates with the lateral 
ethmoid. Anteriorly and ventrally the lachrymal is expanded and overlaps the 
upper lateral region of the premaxilla. The ventral border of the lachrymal is 
slightly irregular in outline whilst the dorsal border is thickened and uniform. The 
infraorbital sensory canal was housed within a tube passing through each of the five 
bones which is incomplete in several places. On the lachrymal the sensory canal 
opened on to the surface of the bone in two places, anteriorly and antero-ventrally. 
The latter opening is in line with a groove on the lateral face of the premaxilla, 
which may have contained an anterior extension of the sensory canal. The lachry- 
mal is thickened along the course of the sensory canal and shows some evidence 



pt.f epo 




Fig. 39. Enchodus marchesettii (Kramberger). Restoration of the skull in lateral view. 



88 UPPER CRETACEOUS TELEOSTS 

postero-dorsally of the remains of a subocular shelf. The infraorbital bones decrease 
in length posteriorly, but not in width, none of the bones bearing any posterior 
expansion. The whole circumorbital series is ornamented with scattered tubercles 
which are most prominent above the course of the infraorbital sensory canal. 

Hyopalatine bones. The general configuration of the hyopalatine bones is visible 
but the more detailed composition is masked by the covering of the dermal elements 
of the jaws, and by the rather poor state of preservation of the hyomandibular. 
The hyomandibular is large and deep with an antero-posteriorly elongated head. 
The posterior edge of the hyomandibular bears the condyle for the articulation of the 
operculum in its upper third. The lowermost two-thirds of the hyomandibular 
support the leading edge of the preoperculum. The lateral face of the hyomandi- 
bular is crossed by a crest which arises at the anterior region of the head and passes 
postero-ventrally in a curve to meet the rear margin about midway down. 

The quadrate condyle is prominent and transversely orientated. Postero- 
dorsally the quadrate is indented, the posterior border of the excavation being lower 
than the remainder of the upper margin. This indentation marks the uppermost 
region of the recess on the medial face of the quadrate which housed the symplectic. 
The symplectic is present as a narrow rod of bone in B.M.N.H. number P. 4748. 

The pterygoid bones are indistinct. The ectopterygoid is a straight, laminate 
bone meeting the quadrate posteriorly and articulating with the palatine anteriorly. 
A single row of prominent teeth is present on the ectopterygoid. The endopterygoid 
can be seen within the orbit close to the parasphenoid where it forms part of the 
roof of the mouth. The metapterygoid is large, overlying the margin of the hyo- 
mandibular, but attaching ventrally to the dorsal edge of the quadrate. 

The palatine is the characteristic tumid bone with an enlarged terminal fang at 90 ° 
to the axis of the bone. Posteriorly the palatine extends dorsally where it articu- 
lates with the lateral ethmoid internal to the lachrymal. Anteriorly the palatine is 
closely associated medially with the mesethmoid and vomerine regions. The head 
of the maxilla is just visible above the premaxilla, anterior to the lachrymal, lying in 
an oblique groove on the lateral face of the palatine. The terminal tooth is laterally 
compressed and bears fine longitudinal striations. 

Dermal upper jaw. The dermal upper jaw is shown in lateral view in Text-figure 
39. It is composed of two bones, the premaxilla and the maxilla. The premaxilla 
is large and has two regions fused at 90 ° to each other. Anteriorly, lying in the same 
plane as the skull-roof, there are the premaxillary pedicels which meet each other in 
the mid-line and rest upon the anterior edge of the mesethmoid. Each pedicel 
contains an elongated oval fenestra within it. The remainder of the premaxilla is 
positioned laterally and is triangular in form, being drawn out antero-posteriorly. 
On the anterior region of the lateral face near to the pedicel there is a groove running 
parallel with the plane of the pedicel. This groove is in line with the anterior opening 
of the supraorbital sensory canal on the frontal. The premaxilla is overlapped 
dorsally by the ventral expansion of the lachrymal bone. Ornamentation in the 
form of scattered tubercles is present on the outer face of the premaxilla, the pedicel 
being more strongly ornamented than the rest of the bone. The premaxilla tapers 



ESPECIALLY MYCTOPHOIDS 89 

posteriorly and terminates just behind the end of the lachrymal, forming the 
anterior half of the oral margin of the upper jaw. There is a single marginal row of 
teeth which are small, acutely pointed and fused directly to the bone with no 
evidence of a tooth base. The teeth are evenly spaced and decrease in size posteriorly 
where they appear to be no more than an extension of the tubercular ornamentation. 
The maxilla is an elongate, uniformly narrow strut of bone, entering the gape 
posteriorly. The head of the maxilla lies in a groove on the dorso-lateral surface of 
the palatine. The anterior region of the maxilla, behind the head, bears a small 
lateral flange under which the postero-dorsal edge of the premaxilla fitted. This 
flange tapers posteriorly to merge with the main shaft of the maxilla behind the 
premaxilla. The ornamentation of the maxilla is in the form of minute tubercles, 
which begin as two rows on the lateral flange and continue back as three rows along 
the remainder of the maxilla. The maxilla expands slightly towards its extreme 
posterior end and is laterally compressed where it rests against the lateral face of the 
mandible. 

Mandible. The mandible is shown in lateral view in Text-figure 39. It is long 
and deep, but the depth is greatly reduced at the symphysis. The dentary is the 
largest component, forming the whole of the upper border, the anterior two-thirds 
of the lower border, and approximately half of the lateral face of the jaw. Internally 
the dentary gives rise to a tooth-bearing flange on the oral margin. This flange 
bears a single row of evenly spaced, relatively large, acutely pointed teeth, confined 
to the anterior two-thirds of the length of the jaw. The teeth are about 8 in number, 
laterally compressed, and decrease in size posteriorly. At the symphysis the 
anteriormost tooth is approximately twice as long as any of the other teeth in the 
row. Immediately in front of this tooth, at the extreme symphysial end, are two 
smaller teeth which are about one-quarter of the length of the longest tooth. All of 
the teeth are, however, identical in other respects. A second row of teeth is present 
on the extreme oral margin of the mandible, composed of between 30 and 40 small, 
pointed, laterally compressed teeth decreasing in size posteriorly. 

The facet on the articular is deeply concave. The anterior edge of the facet is 
rolled upwards around the condyle, whilst posteriorly a short retroarticular process 
curves around behind the condyle. Postero-dorsally the articular is noticeably 
thickened. It joins the posterior end of the oral margin of the dentary to produce 
the coronoid process. The articular extends anteriorly to form the posterior region 
of the lateral face of the mandible. 

The angular is a very small slip of bone on the postero- ventral angle of the jaw, 
below the articulatory facet. 

The mandibular sensory canal ran along the ventro-lateral part of both the 
articular and the dentary. This canal entered the articular through a pore in the 
lateral face of the retroarticular process and passed within a tube of bone beneath 
the articular facet. Anteriorly the sensory canal continued forwards in a shallow 
groove in the dentary. The lateral face of the mandible is ornamented most 
markedly at either end of the jaw. The tubercles on the dentary radiate backwards 
from the symphysis, and on the articular forwards from below the facet. 



go 



1 ITER CRETACEOUS TELEOSTS 



Opercular bones. The opercular apparatus is shown in lateral view in Text-figure 
39. The preoperculum is deep and narrow and curved somewhat forwards ventrally. 
It terminates dorsally below the head of the hyomandibular. The anterior edge of 
the preoperculum is thickened, this thickening is most pronounced in the ventral 
region, where the preoperculum abuts against the rear edge of the crest on the hyo- 
mandibular and the hind edge of the quadrate. The preoperculum widens ventrally 
and is produced into a short backwardly projecting spine. The preopercular 
sensory canal ran in a tube within the bone, opening dorsally by a single pore, and 
ventrally by three openings. Through the most anterior of these ventral openings 
the sensory canal passed into the mandible (B.M.N.H. number P. 9250). The other 
ventral pores are directed more posteriorly. Laterally the preoperculum overlaps 
the anterior opercular region and bears an ornamentation of ridges and scattered 
tubercles, which are more pronounced on the ventral expanded area. 

The operculum is large and deeper than it is long. The facet for the opercular 
condyle of the hyomandibular is in the upper half of the bone. The dorsal and 
posterior margin of the operculum is smoothly convex. From the opercular facet 
on the anterior edge a horizontal strengthening ridge crosses the internal face of the 
operculum, terminating on the posterior edge. The lateral face of the bone is 
ornamented with radiating rows of tubercles. 

The suboperculum is large and deep, its dorsal part lying medial to the ventral 
part of the operculum. Antero-ventrally the bone is more robust and bears a 




10mm 



Fig. 40. Enchodus marchesettii (Kramberger). Pectoral girdle of the left side in 

lateral view. 



ESPECIALLY MYCTOPHOIDS 91 

vertical thickened rib on its anterior edge. Like the operculum, the suboperculum 
is ornamented with radiating tubercles. This radiation originates from a thickened 
knob at the antero-dorsal corner of the bone. The ornamentation is most pro- 
nounced ventrally and is absent where the bone lies medial to the operculum. 

Pectoral girdle and fin. The pectoral girdle is shown in lateral view in Text-figure 
40. The post-temporal has a flattened oval upper limb articulating with the 
epiotic. The dorsal surface of this upper limb is ornamented with tubercles radiating 
anteriorly and medially from a point on the postero-lateral region of the upper 
surface. The ventral limb of the post-temporal is approximately equal in length 
to the dorsal limb but is neither expanded nor flattened, simply being a narrow 
strut of bone articulating with the intercalar. Posteriorly the post-temporal is 
produced behind the junction of the dorsal and ventral limbs forming a facet into 
which the medial face of the head of the supracleithrum fitted. The lateral line 
canal entered the post-temporal bone through a postero-medial foramen in the 
anterior region of the supracleithral facet. It ran within a tube to open anteriorly 
in the angle between the upper and lower limbs of the post-temporal. 

The supracleithrum is a long narrow bone with an expanded head housed in the 
medial facet of the post-temporal. The lateral line canal entered the head of the 
bone postero-laterally and left anteriorly within the post-temporal facet. The main 
shaft of the supracleithrum lies medial to the rear margin of the operculum. 
Ventrally it is attached to the lateral face of the upper extent of the cleithrum. The 
outer face of the bone is ornamented with a few longitudinal striations. 

The cleithrum is large with a curved anterior border which is inclined medially to 
form the posterior limit of the branchial opening. The cleithrum narrows dorsally 
and ends medial to the ventral part of the supracleithrum. Ventrally the cleithrum 
is expanded laterally and terminates beneath the preoperculum. The expanded 
lateral region of the bone is ornamented with the characteristic tubercular ornamen- 
tation, while the more dorsal region is extensively ridged. 

The endoskeletal girdle is masked by the lateral cleithral expansion. It does, 
however, project slightly, posterior to the cleithrum, where a scapula and coracoid 
are visible as well as a prominent scapular foramen. A saddle-shaped condyle on 
the rear edge of the scapula provided the articulatory point for the first fin ray. 

The pectoral fin contains approximately 13 rays, all of which are segmented and 
branched distally except for the first ray which does not appear to be branched. A 
small postcleithrum is attached to the posterior edge of the ventral part of the 
supracleithrum. 

Pelvic girdle and fin. The pelvic bone of either side is elongated, narrow and 
tapers to a point anteriorly, but is expanded posteriorly to produce the articulatory 
region for the pelvic fin rays. Medial to this articulatory region each pelvic bone 
bears a prominent horizontal flange which extends into the mid-line to contact its 
partner from the opposite side. The articulatory region consists of a projecting 
lateral condyle, supported by a prominent ridge. The rays of the pelvic fin are all 
segmented and branched distally except for the first ray which does not appear to 



92 



I'IM'ER CRETACEOUS TELEOSTS 




o 




■p 




_o 








o 




# 




en 




to 


CJ 


A 





+-> 


-^ 


m 


-7 


O 


O 


c 


ft 





fl 


+j 


' r ~ l 


rt 


en 


M 


4) 


o 


-t-^ 


+J 


3 


t/> 


O 


V 


C/3 


ft 






rt 




U) 




|H 




o 




TjJ 


be 


-d 




1 


a 


4-" 


rt 


3 


h 


,a 


w 


d 


"" — 


o 




•M 


•fi 


+J 






"^ 


"a 




o 


-^ 


w 


Vi 
X 


_i; 


« 


- 




o 


^ 


tn 


to 


<L> 




^C 


o 


1 — ' 


-^ 




o 




^ 




^ 




M 




* 




6 








fc 





ESPECIALLY M YCTOPHOIDS 93 

be branched. The number of rays contained in the fin is approximately 10. The 
position of the pelvic fins is abdominal, below the origin of the dorsal fin. 

Vertebral column. The vertebral column is shown in the restoration in Text-figure 
41. It consists of 37 vertebrae, of which 24 are caudal. The centra are slightly 
constricted mesially and are just longer than deep. The lateral face of each centrum 
bears a pronounced longitudinal ridge above and below which are prominent 
depressions. The neural arches occupy the entire length of the dorsal region of each 
centrum, and the neural spines emerge from the mid-dorsal region of each arch. 
The suture between the neural arches and the centra is always clearly visible. 
Anteriorly the neural arches have invariably become separated from the correspond- 
ing centra during fossilization indicating the looseness of this connection. More 
posteriorly the neural arches tend to be more closely attached to the centra. The 
precaudal vertebrae bear paired ventro-lateral transverse processes near the anterior 
end of each centrum. Pleural ribs articulate with these processes, the ribs them- 
selves being long and flattened, and tapering to a point. The first two or three 
caudal vertebrae have a haemal arch with a gradually lengthening haemal spine, but 
also bear articulated pleural ribs of reduced dimensions. The subsequent haemal 
arches bear longer backwardly directed haemal spines. 

Zygapophyses are present both above and below the column. The postzyga- 
pophyses are the more prominent dorsally, and the anterior zygapophyses are more 
prominent ventrally. 

Epineurals and epipleurals are associated with the first 20 vertebrae. 

Median fins and tail. The dorsal fin is positioned midway along the back and is 
composed of 16 rays, all segmented and all but the first branched. The first ray is 
shorter and stouter than the subsequent rays, the second is the longest. The 
rays decrease in size posteriorly. Each fin ray is supported by a pterygiophore, 
comprising an elongated proximal radial and a medial radial, the latter extending at 
a slightly oblique angle between the bases of adjacent fin rays. The distal radials 
are not distinguishable as separate elements, they are presumably incorporated 
into the fin ray bases. The more anterior proximal radials are expanded, especially 
the first which bears a large median keel projecting anteriorly. 

The anal fin is composed of 20 rays and is confined to the more posterior part of 
the body. As in the dorsal fin the first ray is short and unbranched, as opposed to 
the longer, branched and segmented, subsequent rays. The radial composition is 
like that of the dorsal fin except that the most anterior proximal radial is unexpanded 
and the medial radials are less pronounced. 

The caudal skeleton (Text-fig. 42) is composed of 6 vertebrae : three separate 
preural vertebrae (numbers 2, 3 and 4) ; a compound vertebra formed from the 
combination of preural vertebra 1 with ural vertebra 1 ; and a terminal half-centrum, 
the remains of ural vertebra 2. Preural vertebrae 3 and 4 bear slightly expanded, 
backwardly projecting neural and haemal spines which aid in the support of the 
accessory fin rays. The second preural vertebra (the first free preural vertebra) 
does not possess a neural spine, simply an expanded neural arch dorsally, and a 
markedly more expanded haemal spine ventrally. Preural vertebra 1 and ural 



94 



UPPER CRETACEOUS TELEOSTS 
ur2 

Sl j n.a. pu2 

n.s 



pul+ul 




5mm 



Fig. 42. Enchodus marchesettii (Kramberger). Caudal fin skeleton in lateral view. 

vertebra 1 are fused with one another to produce a compound centrum which is 
slightly upturned. Articulating with this compound centrum are three structures, 
the parhypural, which represents the haemal spine of the first preural vertebra, 
and two hypural bones which extend up to the mid-line. The first hypural is the 
larger of the two, and basally bears a prominent fenestra. Dorsally the compound 
centrum bears an enlarged stegural element which projects postero-dorsally. The 
stegural bears an antero-dorsal expansion extending forwards towards the neural 
spine of the third preural vertebra. Posterior to the compound centrum ural 
vertebra 2 bears four hypurals, and may bear a second uroneural behind the postero- 
dorsal shaft of the stegural on the first ural vertebra. 

The caudal fin consists of 19 principal rays, 17 of which are branched. There 
are 10 shorter accessory rays dorsally and ventrally in advance of the principal 
caudal rays. 

Scales. The body is devoid of scales except for a row along the course of the 
lateral line and three reduced dermal scutes between the occiput and the dorsal fin. 
The line of scales marking the passage of the lateral line begins at the posterior region 
of the head of the supracleithrum, and extends along the whole length of the body, 
terminating on the lateral surface of the caudal peduncle. The row consists of 37 
scales, the same number of scales as vertebrae, and each one is roughly rectangular 
in shape with its posterior margin deeply excavated. Each individual scale is very 
small and is isolated from its neighbours. The surface of each scale bears an 
ornamentation of small ridges and a backwardly directed spine which projects over 
the anterior region of the deep indentation. The spine is very small on the anterior 



ESPECIALLY M YCTOPHOIDS 95 

scales of the row but becomes progressively more pronounced posteriorly until, on 
the tail peduncle, it is an extremely prominent flange. 

Genus PALAEOLYCUS von der Marck, 1863 

Diagnosis (emended) : Enchodontidae in which the body is elongated and no 
deeper than the depth of the head at the occiput. Vertebral number approximately 
73 of which 40 are caudal. Pectoral fins greatly enlarged, pel vies excessively 
reduced. Anal fin greatly extended with 50 rays occupying one-third of the total 
body length, its origin being in advance of the dorsal fin. Dorsal fin itself much 
reduced in extent and positioned midway along the body. Mid-dorsal scutes not 
greatly reduced, 5 or 6 in number and almost overlapping. 

Type species. Palaeolycus dreginensis von der Marck 

Palaeolycus dreginensis von der Marck 

(Text-fig. 43) 

1863 Palaeolycus dreginensis von der Marck : 31, pi. 4, fig. 7. 
1954 Palaeolycus dreginensis von der Marck ; Siegfried : 13, pi. 5, figs. 1, 2, 2a. 
Siegfried lists earlier references. 

Diagnosis. As for genus, only species. 

Holotype. Number 808 in the Palaontologisch.es Institut der Westfalische 
Wilhelms-Universitat, Miinster, from the Upper Senonian of Sendenhorst, West- 
phalia. 

Material. The holotype and specimen number 8438, both from Sendenhorst 
and both in Miinster. 

Remarks. Of the two specimens examined, the second specimen, number 
8438, was the more complete and was the specimen used by von der Marck in his 
second work (1885 : 251). Neither specimen is complete by any means so that the 
following description is deficient in many features. No other specimens of the 
species are known to the author. 

Description. Neurocranium. Little of the neurocranium is preserved, and what 
is observable is merely an impression in the matrix. The frontals appear to occupy 
a large proportion of the roof of the skull and are produced laterally as the upper 
border of the orbit. The sphenotic projects ventro-laterally from beneath the 
frontal at the hind end of the orbit and presumably the neurocranium attains its 
maximum width between the extremities of the sphenotics. The postero-lateral 
neurocranial region inclines ventrally and the presence of a vertical crest of bone is 
indicated in the pterotic region. The crest demarcates the dilatator fossa laterally 
and the post-temporal fossa medially, both fossae being unroofed. The neurocranial 
bones are unornamented except for the presence of raised ridges of bone running in a 
longitudinal direction above the orbit. The extent of the cranial cavity region is 
reduced, and is shorter than the diameter of the orbit which itself occupies at least 



96 I'I'PER CRETACEOUS TELEOSTS 

half the length of the neurocranium. The lateral ethmoid, at the anterior end of 
the orbit, is attached to the underside of the frontal, and ventrally is associated with 
a postero-dorsal expansion of the palatine. The frontal tapers anteriorly towards 
the mesethmoid region. 

Jaws and suspensorium. The hyomandibular is large and vertical with part of its 
lateral face covered by the metapterygoid. The quadrate is large and triangular 
with a prominent transverse condyle at its ventral extremity. The ectopterygoid 
is an elongate laminar bone bearing a uniform row of teeth. The teeth are large, 
laterally compressed, just noticeably recurved, regularly spaced and decreasing 
in length towards the posterior end of the bone. The impression of a laminate 
endopterygoid is visible immediately below the orbit. The palatine is large and 
tumid and bears a single, enlarged terminal tooth set at 90 ° to the main body of the 
bone. Posteriorly the palatine is associated with the lateral ethmoid dorsally and 
the ectopterygoid ventrally. The palatine tooth is considerably larger than the 
teeth present on the ectopterygoid but in all other respects closely resembles them. 
The lateral face of the palatine bears traces of an obliquely inclined groove which may 
have housed the head of the maxilla. 

The premaxilla appears large and irregularly triangular in outline. Impressions 
of the oral edge of the bone, between the anterior ectopterygoid tooth and the 
palatine tooth, show clearly that the premaxilla bore a row of small acutely pointed 
teeth. The premaxilla may occupy more than half of the oral border of the upper 
jaw with the maxilla entering the gape behind it. The maxilla is a stout, smooth, 
untoothed strut of bone. 

The mandible is long and deep, its maximum length being almost four times its 
depth, and its minimum depth occurring at the symphysial end. The dentary 
forms the complete upper border of the jaw and well over half of the lateral face and 
the ventral margin. Teeth are present in a regularly spaced row with the anterior 
tooth at the symphysial end being greatly elongated. The jaws are closed in the 
specimens examined and the anterior dentary tooth lies against the anterior face of 
the palatine tooth and protrudes dorsal to it. The rest of the mandibular teeth 
are at most half as long as the anterior tooth and they decrease in size posteriorly. 
The articular bone forms the posterior end of the jaw and bears the articular facet 
for the reception of the condyle. The facet is rolled both anteriorly and posteriorly 
around the condyle and a small retroarticular process is produced posteriorly. 

The preoperculum could not be observed in either of the specimens examined, 
although from the uprightness of the suspensorium it can be assumed that the 
preoperculum was vertical and relatively narrow. The operculum is large and 
considerably deeper than it is long. The facet for the articulation with the hyo- 
mandibular is positioned approximately midway along the straight anterior edge of 
the bone. The posterior edge of the operculum is smoothly rounded throughout. 
A horizontal strengthening ridge passes across the inner face of the bone from the 
articulatory facet to the posterior edge. The lateral face of the operculum is 
ornamented with a few weak ridges radiating from the region of the articulatory 
facet. A suboperculum is present but its limits are difficult to determine. 



ESPECIALLY MYCTOPHOIDS 





97 



a 
o 

Id 
M 

en 

(D 
^ d 

*" 5 

O +J 

.2 ft 
rt 



« 5 



^ 


en 


o 


Ih 


u 


O 


ni 


TD 


s 


X) 


l-l 
03 


"a 


-o 






+j 


a 





o 


X 


> 


T3 


<o 


0) 


■*» 


-4-> 


eo 


■P 


s 






1 

o 




in 

5 


ta 


d 


«o 


o 


X 


en 


Vi 


oj 


-o 


a 


o 




■a 




-2 








« 




&, 




en 




•*■ 




6 








h 





98 UPPER CRETACEOUS TELEOSTS 

Paired fins, median fins and tail. A post-temporal bone of modest dimensions lies 
above the operculum, posterior to the occiput, and is composed of two limbs joined 
posteriorly. The dorsal limb is the more robust of the two. The supracleithrum is 
absent in both of the specimens and the cleithrum is also poorly preserved. The 
pectoral fin is enormous, arising well up on the flank while below its base a slight 
backward projection of the coracoid is visible. The fin consists of approximately 
14 rays and is as long as, if not longer than, the entire length of the head (without 
the operculum). 

The pelvic fin is much reduced and positioned about midway between the origin 
of the pectoral fin and the beginning of the anal fin. The exact number of pelvic 
fin rays could not be ascertained but it is obviously small. The rays themselves 
are very feeble. 

The very small dorsal fin arises behind the origin of the anal fin, and appears 
to consist of 9 rays. The first ray is shorter than the subsequent ones and is un- 
branched. The second ray is the longest and the remainder decrease in length 
posteriorly. All of the fin rays are segmented and most, if not all, are branched. 

The anal fin is greatly extended, its origin being a little over half way back along 
the body while it terminates close to the caudal fin. Forty-seven rays are visible in 
the fin, but there may possibly be a few more present (50) . As in the dorsal fin the 
first ray is shorter than the rest and is unbranched. 

The caudal fin is large, deeply cleft and consists of 19 principal rays, 17 of which are 
branched and segmented. Shorter, unbranched accessory rays are present both 
epaxially and hypaxially, there being 14 in all. The composition of the caudal 
skeleton was undeterminable. 

Vertebral cohimn. The body is considerably elongated and there appear to be 73 
vertebrae, of which approximately 33 are precaudal. The individual centra are 
mesially constricted and longer than they are deep, although at the posterior end of 
the column the length of each centrum is reduced so that they present practically a 
square outline. Each centrum bears a neural arch and a neural spine, the latter 
arising from the middle of each of the neural arches. In the anterior precaudal 
region the neural spines are straight, but more posteriorly the spines show a marked 
backward curvature. The presence of a persistent notochord is shown by a con- 
tinuous perforation throughout the length of each centrum. The precaudal verte- 
brae bear fine elongated pleural ribs which taper gradually to points. Haemal 
arches and spines characterize the caudal region, the arches arising from the anterior 
region of each centrum. Intermuscular bones, both epineurals dorsally and epi- 
pleurals ventrally, occur in continuous series along practically the entire length of 
the body. 

Scales. The body is devoid of scales except for a row of dorsal scutes in the mid- 
line behind the occiput. There may be as many as 5 or 6 of these scutes and each 
one is ornamented with radiating bony ridges. No traces of lateral line scales were 
found in either specimen. 



ESPECIALLY MYCTOPHOIDS 



99 



Family EURYPHOLIDAE nov. 

Diagnosis. Head not deepened and the body shallower than the greatest depth 
of the head which is at the occiput. Post-temporal fossa roofed. Lower jaw long 
and relatively shallow, the symphysis only slightly constricted. Articular facet not 
visible from the lateral aspect, being covered by a flange of bone laterally. Oper- 
culum as wide as deep with a posterior spine ; the preoperculum also bears a 
prominent posterior spine ventrally. Pelvic fins as large as or larger than pectoral 
fins. Pectoral fins raised slightly on the flanks ; pelvics sub-thoracic or abdominal 
in position. Cleithrum expanded posteriorly below the fin insertion. Lateral line 
scales overlap. Mid-dorsal scutes large and overlapping. 



Genus EURYPHOLIS Pictet, 1850 

Diagnosis (emended). Eurypholidae in which the head is not particularly 
elongate and not shallow. Vertebral number between 35 and 40 of which the caudal 
region remains constant at 20 vertebrae. Body not lengthened or reduced in depth, 
maximum depth equalling the depth of the head at the occiput. Pelvic fins larger 
than the pectoral fins and sub-thoracic in position. Mid-dorsal scutes three in 
number and overlapping. 

Type species. Eurypholis boissieri Pictet. 

Remarks. The genus Eurypholis is known by only two species, the type species 
and one other which was originally assigned to the genus Enchodus by Woodward 
(1901 : 193, Enchodus pulchellus). 

Eurypholis boissieri Pictet 
(Text-figs. 44-48) 



1850 Eurypholis boissieri Pictet 

1866 Eurypholis boissieri Pictet 

1887 Eurypholis boissieri Pictet 

1 90 1 Eurypholis boissieri Pictet 

1902 Eurypholis boissieri Pictet 
1926 Eurypholis boissieri Pictet 
1930 Eurypholis boissieri Pictet 
1932 Eurypholis boissieri Pictet 
1940 Eurypholis boissieri Pictet 
1958 Eurypholis boissieri Pictet 
1964 Eurypholis boissieri Pictet 
1966 



30, pi. 4, figs. 2-4. 

Pictet & Humbert : 102, pis. 15, 16. 

Davis : 596. 

Woodward : 207, pi. 11, fig. 7, text-fig. 7. 

Woodward : 56, fig. 12. 

Woodward : 203, fig. 4. 

Maury : 96, pi. 1, fig. 2. 

Woodward : 160, fig. 247. 

Berg : fig. 143. 

Bertin & Arambourg : figs. 14, 16. 

Danil'chenko : 412, fig. 128. 

Lehman : 203, figs. 204, 205. 



Eurypholis boissieri Pictet 
Diagnosis (emended). Eurypholis of standard length not exceeding 10 cm. 
Head with opercular apparatus about twice as long as deep and occupying one-third 
of the standard length. Greatest depth of the mandible equal to one-sixth of the 
total length. Maxilla finely toothed. Dentary bearing two rows of teeth. Pre- 
operculum deep, narrow and vertical. Operculum as deep as broad. Dorsal fin 
with 13 rays; anal fin with 14 rays arising behind the dorsal, nearer to the caudal 
peduncle than to the pelvic fin. 



ioo UPPERCRETACEOUS TELEOSTS 

Holotype. Specimen number 639/58 in the Museum d'Histoire Naturelle, 
Geneva, from the Middle Cenomanian of Hakel, Lebanon, a head with the precaudal 
region. 

Material. The holotype in Geneva and fifteen specimens in the B.M.N.H. The 
specimens used by Pictet (1850) and Pictet & Humbert (1866) were also examined 
in Geneva. All of the material examined was from the two Middle Cenomanian 
localities in the Lebanon, Hakel and Hajula. 

Remarks. All of the B.M.N.H. specimens were prepared in acetic acid by the 
transfer method. The only part of the skeleton which remains incompletely 
described is the neurocranium. All of the specimens of Enrypholis recovered from 
the Lebanese deposits are referable to the type species, being identical in body 
proportions and fin ray counts. The only variable feature in the osteology is the 
vertebral number, but this variation is within very narrow limits and the variability 
is confined to the precaudal region, the caudal region remaining constant at 20 
vertebrae. 

The references in the synonymy which occur after 1901 are simply mentions of the 
specific name together with the reproduction of Woodward's reconstruction of the 
body (1901 : text-fig. 7). 

Description. Neurocranium. The neurocranium is shown in dorsal view in 
Text-figure 44. The cranial roof is flattened and widest at the hind border of the 
orbit between the sphenotics, and almost entirely composed of the frontals which 
extend back to near the occipital border. There is a straight suture between the 
frontals and a shallow median depression between the orbits which is more sparsely 
ornamented than the remainder of the skull-roof. The frontal extends on to the 
dorso-medial surface of the sphenotic, and the sphenotic just projects laterally from 
under it. The roof is covered by an ornamentation of raised bony tubercles 
radiating in all directions. Above the orbit and along the course of the supra- 
orbital sensory canal the tubercles are present on raised ridges of bone. Anteriorly 
the frontal tapers to end by insertion into (or under) the rear edge of the short, 
broad mesethmoid. 

The supraoccipital is small, meeting the frontals anteriorly and separating the 
parietals laterally. In the mid-line the supraoccipital bears a small backwardly 
directed crest which does not extend dorsally above the plane of the skull-roof. 

The parietals are transversally orientated, relatively narrow strips of bone bordered 
anteriorly by the frontals and laterally by the pterotics. Both the parietals and the 
supraoccipital are in part covered by the backward extension of the frontals. The 
tubercular ornamentation of the frontals is continued on the parietals and the supra- 
occipital. Near the posterior edge of the parietal a low transverse crest is present 
which functioned as an area for muscle insertion. The narrow strip of parietal and 
that portion of epiotic which is visible behind the crest are devoid of ornamentation. 

The pterotic forms the postero-lateral border of the skull-roof, contacting the 
sphenotic anteriorly and the frontal and parietal medially. The dorsal surface of 
the pterotic is ornamented with tubercles which are connected by ridges of bone 



ESPECIALLY M YCTOPHOIDS 



giving an irregular latticed effect. This joining up of the tubercular ornamentation 
is visible to a certain extent on the postero-lateral region of each frontal. 



spo 




5 mm 



infs.c 



pop.s.c 



med.s.c 



Fig. 44. Eurypholis boissieri Pictet. Neurocranium in dorsal view. The broken line 
on the right side of the figure indicates the course of the sensory canals. 



102 UPPER CRETACEOUS TELEOSTS 

The sphenotic forms the posterolateral corner of the orbit, meeting the pterotic 
behind and the frontal medially. The dorsal surface of the sphenotic is in part 
ornamented with the latticed tubercular ornamentation already noticed on the 
pterotic. 

The course of the sensory canals on the skull-roof is indicated by tubes within the 
bone and by grooves and depressions on the surface. The supraorbital sensory canal 
ran within a tube in the frontal opening anteriorly on the edge of the bone just 
posterior and lateral to the mesethmoid. Posteriorly this tube opens above the 
sphenotic in a triangular depression which is devoid of ornamentation. The main 
branch of the infraorbital sensory canal passed on to the dorsal surface of the 
sphenotic and entered a short tube within that bone which opens medially in the 
depression into which the supraorbital sensory canal passed. The otic branch of the 
infraorbital sensory canal entered the pterotic posteriorly through a pore at the 
anterior end of a short open groove. The sensory canal passed anteriorly within the 
pterotic and opened on to the dorsal surface of the sphenotic. Laterally, near 
the posterior end of the pterotic, a further foramen is present through which the 
preopercular sensory canal passed. A supratemporal canal was present running 
transversely across the skull-roof. This sensory canal connected with the otic 
branch of the infraorbital canal within the open groove on the posterior part of the 
pterotic. The canal passed medially in a tube within the parietal, but about mid- 
way along the bone the tube opens into a groove. This open groove is continuous 
across the dorsal surface of the supraoccipital from one parietal to the other. A 
further groove is seen in the mid-line of the supraoccipital connecting with and 
projecting back from the transverse groove. This groove seems to indicate that a 
further sensory canal was present in the mid-dorsal line of the body which connected 
with the supratemporal canal. 

The mesethmoid is a short broad bone which is well ossified and closely attached 
to the dorsal surface of the vomer. The palatines articulate with the lateral wings of 
the mesethmoid, while the premaxillary pedicels are joined in the mid-line and 
articulate with the anterior edge of the mesethmoid. 

Infraorbital bones. The infraorbitals are shown in lateral view in Text-figure 45. 
There are four infraorbitals and a large lachrymal in the circumorbital ring of bones. 
The bones decrease in size posteriorly and are unexpanded. Each bone in the ring 
carries a tube through which the infraorbital sensory canal ran. This tube opens 
occasionally on to the surface where it is simply a groove roofed by a flange of bone. 
The region of the sensory canal is ornamented with a single row of tubercles and the 
remainder of each infraorbital bone is sparsely tuberculated. The posterior bone of 
the series articulates with the sphenotic and received the sensory canal from the 
latter. The lachrymal is the anterior bone of the series and is longer than deep and 
dorsally bears a subocular shelf, but the remainder of the bones do not appear to 
exhibit this shelf. From the lachrymal an antero-dorsal projection articulates with 
the lateral ethmoid, while an anterior projection overlies the premaxilla laterally. 
The ventral border of the lachrymal is rounded and slightly roughened. The tube 
which contained the sensory canal runs within a dorsal ornamented thickened region, 



ESPECIALLY MYCTOPHOIDS 103 

and opens in two places on to the surface of the bone. One opening faces anteriorly, 
the other ventrally. The branch of the sensory canal which passed out through the 
anterior opening possibly continued on the lateral surface of the premaxilla where a 
distinct groove is observable. 

Hyopalatine bones. The hyomandibular is large and has an elongated head. 
The lateral face bears a crest originating on the anterior region of the head and 
passing postero- ventrally to the rear of the bone where it is joined by a further crest 
running ventrally from the posterior region of the head. The combined crest passes 
ventrally, and shallows until it disappears at the ventral extremity of the bone. The 
opercular condyle is found on the posterior edge of the hyomandibular within the 
upper half of the bone. Laterally, anterior to the crest, the hyomandibular is thin 
and covered in part by the expanded metapterygoid. The quadrate condyle is 
normally not visible from the lateral aspect since it is covered by an upgrowth of the 
articular lateral to the articular facet. The quadrate is triangular in shape, and 
indented postero-dorsally at the upper limit of the groove on the medial face which 
contained the symplectic. The pterygoid region bears a uniform series of large, 
laterally compressed, acutely pointed teeth. The palatine bone is prominent and 
tumid and attached to the anterior end of the pterygoids. Postero-dorsally the 
palatine is slung from the lateral ethmoid and antero-medially it appears to be in 
association with the mesethmoid region. The palatine bears one enlarged terminal 
tooth which is set at 90 ° to the axis of the bone. This tooth is larger than those of 
the ectopterygoid but of similar form. 

Dermal upper jaw. The dermal upper jaw is shown in lateral view in Text-figure 
45. It consists of two bones, the premaxilla and the maxilla. The premaxilla is 
large and in two main regions which are at an angle of 90 ° to each other. Anteriorly 
there is a premaxillary pedicel in the same plane as the skull-roof, containing 
an elongated fenestra. The pedicels meet in the mid-line and articulate with 
the anterior edge of the mesethmoid. The second region of the premaxilla is 
lateral in position and fuses with the lateral edge of the pedicel. In the region of 
fusion between the two parts of the premaxilla there is a deep groove passing anter- 
iorly to the edge of the snout region. This groove is in a direct line with the anterior 
opening of the supraorbital sensory canal on the antero-lateral edge of the frontal. 
The postero-dorsal region of the premaxilla is overlapped laterally by the anterior 
region of the lachrymal. A ventro-lateral groove passing anteriorly to the edge of the 
snout possibly contained the most anterior region of the infraorbital sensory canal. 
The entire bone is ornamented with tubercles except in the mid-line on the pedicel. 
The premaxilla decreases in depth beneath the lachrymal and tapers posteriorly. 

The maxilla enters the gape behind the premaxilla, the margins of both bones 
forming a continuous edge to the oral border. The maxilla is fused to the premaxilla 
and the line of demarcation is difficult to define. The maxilla has no ornamentation. 
The oral border bears a marginal row of teeth continuous with those on the pre- 
maxilla. The teeth are small, evenly spaced, acutely pointed and about 25 in number. 
Mandible. The mandible is shown in lateral view in Text-figure 45. It is long 
and shallow, the depth being somewhat reduced in the symphysial region. The 



104 



asppm 



UPPER CRETACEOUS TELEOSTS 

fr 




pops 



Fig. 45. Eurypholis boissieri Pictet. Restoration of the skull in lateral view. 

dentary is the largest of the three bones and forms the complete upper border, the 
majority of the lower border and the anterior two-thirds of the lateral face. The 
posterior edge of the dentary is V-shaped. Internally the dentary gives rise to a 
tooth bearing flange on the oral margin which supports a single series of evenly 
spaced teeth, about 10 in number. The teeth are pointed, laterally compressed, 
slightly recurved, hollow and with expanded bases. The teeth are of approximately 
the same size except for the penultimate tooth at the symphysial end of the mandible. 
This tooth is about twice the length of the other teeth in the row, but in all other 
respects closely resembles them. A second row of teeth on the oral margin of the 
mandible is composed of between 40 and 50 small, pointed, laterally compressed teeth 
which decrease in size backwards. In the posterior region these small teeth merge 
imperceptibly with the tubercular ornamentation on the lateral face of the mandible. 

The articular facet is hidden from external view by a vertical upgrowth of bone 
from the lateral face of the articular. The articular forms, laterally, the posterior 
third of the mandible, occluding the V-shaped indentation of the dentary. 

The angular is a small slip of bone forming the postero-ventral angle of the mand- 
ible beneath the articular facet. Its posterior face is slightly concave. 

The mandibular sensory canal traversed both the articular and the dentary 
ventro-laterally. The lateral face of the mandible is ornamented with numerous 
tubercles, which are more prominent at the symphysis where they are raised on to 
ridges. The tubercles radiate in lines back from the symphysis on the dentary, and 
on the articular upwards and forwards from below the articular facet. 

Opercular bones. The opercular bones are shown in lateral view in Text-figure 45. 
The preoperculum is deep and narrow and curved slightly forwards. It terminates 



ESPECIALLY MYCTOPHOIDS 105 

dorsally near to the uppermost region of the hyomandibular, and has a thickened 
anterior edge lying against the posterior edge of the hyomandibular crest. Ventrally 
the thickened edge of the preoperculum rests in a groove on the posterior thickened 
edge of the quadrate and is drawn out posteriorly into a prominent pointed spine. 
The preopercular sensory canal entered the bone through a dorsal pore and ran in a 
tube along the length of the bone, opening ventrally by two or three pores on or 
adjacent to the ventral spine. All of the visible openings are directed posteriorly, 
but the opening which transmitted the sensory canal to the mandible is usually 
hidden beneath the postero-dorsal upgrowth of the articular. The preoperculum is 
ornamented with the characteristic tubercles which are more prominent over the 
sensory canal. Posteriorly the preoperculum is thin and sparsely ornamented, 
although ventrally on the spine the tubercles are present on ridges of bone. 

The operculum has a slightly concave anterior edge which abuts against the 
posterior face of the hyomandibular. Dorsally the operculum is rounded but 
ventrally and postero-ventrally the edge is slightly concave. From the opercular 
facet on the anterior edge a horizontal strengthening ridge is produced which crosses 
the internal face of the bone to terminate on the lower posterior angle where it is 
extended as a short spine. The lateral face of the operculum is ornamented with 
bony tubercles which radiate out from the point of articulation with the hyo- 
mandibular. 

The suboperculum is large and dorsally lies internal to the ventral part of the 
operculum. Ventrally the suboperculum is more robust and bears tuberculated 
ridges, the dorsal region being more sparsely ornamented. 

Branchiostegal rays. There appear to be 14 or 15 branchiostegal rays, the first 8 
or 9 of which articulate with the ventro-lateral face of the ceratohyal, the tenth 
articulates with the region between the ceratohyal and epihyal, whilst the remaining 
4 or 5 articulate with the lateral face of the epihyal. The anteriormost rays, on the 
ceratohyal, are smaller and thinner than the subsequent ones, which are broader 
based, more widely separated, and taper more rapidly. The terminal branchiostegal 
ray is the broadest of the whole series and arises on a level with the ventral region of 
the preoperculum, curving postero-dorsally along the ventral margin of the sub- 
operculum. 

Pectoral girdle and fin. The pectoral girdle is shown in lateral view in Text-figure 
46. The post-temporal is horizontally arranged with an oval flattened upper limb 
articulating anteriorly with the epiotic. The dorsal surface of this upper limb is 
densely ornamented with raised bony tubercles. The ventral limb of the post- 
temporal is an elongated strut of bone, equal in length to the upper, which articu- 
lates antero- ventrally with the intercalar. The two limbs of the post-temporal join 
and the bone is produced posteriorly as a small vertical flange. A postero-medial 
facet is found on this flange and the facet contained the anterior region of the head 
of the supracleithrum. The lateral line canal passed within a tube through the bone. 
The anterior opening of the tube is clearly shown on the lateral face of the post- 
temporal. 



inii 



UPPER CRETACEOUS TELEOSTS 



The supracleithrum is long and narrow, with an antero-posteriorly elongated head. 
The lateral line canal passed straight through the head of the supracleithrum within 
a tube. Ventrally the supracleithrum lies against the upper lateral face of the 
cleithrum, and here the supracleithrum bears several vertical striations on its 
lateral face. 




5mm 



Fig. 46. Eurypholis boissieri Pictet. Pectoral girdle of the right side in lateral view. 



The cleithrum (post-clavicular plate of Woodward, 1901 : 208, fig. 7) is a very 
large and characteristically shaped bone. The anterior margin is inclined medially 
where it forms the posterior boundary of the branchial opening, and the lateral 
face is greatly expanded both anteriorly and posteriorly, particularly in the ventral 
region. The anterior termination of the cleithrum lies medial to the extreme 
ventral end of the preoperculum, and the posterior end of the cleithrum extends for a 
considerable distance backwards below the pectoral fin. Both of these extremities 
are acutely pointed. The ventral region of the cleithrum, as well as being expanded 
anteriorly and posteriorly, is also inflected strongly towards the mid-ventral line of 
the body, approaching its partner from the opposite side. The whole external face 
of the cleithrum is ornamented with the characteristic tubercles, which are most 
prominent near the anterior and the ventral edges. Dorsally the cleithrum tapers 
and lies medial to the supracleithrum. 

The endoskeletal girdle attaches to the internal face of the cleithrum, but owing 
to the great expansion of the cleithrum its outlines are not determinable: a large 



ESPECIALLY MYCTOPHOIDS 107 

scapular foramen is, however, visible. Four small hour-glass shaped radial bones 
are present on the posterior edge of the scapulocoracoid and with these the 13 rays 
of the pectoral fin articulated. The first ray is the longest, the rest gradually decreas- 
ing in size. Due to the posterior expansion of the cleithrum, the pectoral fin is raised 
slightly so that its point of insertion has moved up the flank. 

A small, narrow and curved postcleithral bone attaches to the postero-ventral part 
of the supracleithrum and projects postero-ventrally, lying medial to the radial 
elements of the pectoral fin. 

Pelvic girdle and fin. Each pelvic bone is composed of a roughly triangular sheet 
of thin bone which is flattened and meets its partner in the mid- ventral line just 
posterior to the cleithrum. The pelvic bones therefore can be said to be sub-thoracic 
in position. Each bone bears a thickened condylar region posteriorly which lies 
lateral to an articulatory facet. The condyle is supported by a ridge passing 
anteriorly along the lateral edge of the bone. The pelvic fin consists of 8 stout rays 
all of which are branched distally and segmented. 

Median fins and tail. The median fins are shown in the restoration, Text-figure 47. 
The dorsal fin is somewhat deeper than long and contains 12 rays. The first ray is 
shorter than the subsequent ones and segmented but not branched. The elongated 
proximal radials are prominent and all are expanded to a certain extent, especially 
the first which bears a large median keel projecting anteriorly. The distal ends of 
the proximal radials are expanded at the bases of the fin rays and the medial radials 
extend obliquely between the bases of adjacent fin rays. 

The anal fin consists of 13 or 14 rays. The first ray is shorter than the second and 
is unbranched : the remainder are all branched. All the rays are segmented. The 
pterygiophore composition of the anal fin parallels that of the dorsal fin except that 
the proximal radial at the anterior end is not expanded. 

There are six vertebrae entering into the composition of the caudal skeleton 
(Text-fig. 48), three free preural vertebrae, one free ural vertebra, and a compound 
vertebra formed from the fusion of the first preural vertebra and the first ural 
vertebra. Preural vertebrae 3 and 4 both bear normal neural spines, which are 
strongly inclined posteriorly, and ventrally these preurals bear expanded haemal 
spines. Preural vertebra 2 (the first free preural vertebra) , bears a greatly expanded 
haemal spine ventrally but does not possess a neural spine, simply an enlarged neural 
arch. The compound vertebra formed from the fusion of preural vertebra 1 with 
ural vertebra 1 bears a parhypural ventrally (the haemal spine of the first preural 
vertebra), and two hypurals which extend dorsally to the mid-line. The ventral- 
most hypural (hypural 1) is by far the larger of the two and has an oval foramen 
near to its base. Dorsally the compound centrum bears an expanded stegural 
which extends as a rod-like structure in a postero-dorsal direction, and has a thin 
anterior projection in the mid-line which extends forwards towards the neural 
arch of preural vertebra 2. The second ural vertebra is in the form of a small half- 
centrum which bears a posterior fan of four hypurals supporting the upper lobe 
of the caudal fin. The second ural vertebra also appears to bear a second uroneural 
element closely associated with the posterior edge of the stegural. At least one 



io8 



11'1'ER CRETACEOUS TELEOSTS 




n! 




o 




co 




CD 




C 








. — < 








a 




u 




4> 




-" 




ca 




1 1 








a 




o 




+j 




0) 


d 


"a3 


a 


CO 


'-P 




"co 


4) 


O 


rd 


Cl 


+J 




-+H 


a 


o 


* rH 


a 


cu 





4-1 


1-H 


3 
O 
CO 


o 




+J 


ai 


CO 


CO 


cu 


1-H 


tf 


c 
T3 




X3 








E 


-M 




O 


+j 


ft 






13 


<0 


CO 


■^ 


+J 


<o 


+J 


."> 




o 


1 


<i 


O 


CO 




-C! 




o 




>« 




«*. 




?N 




^ 




s 




m 




t^. 




■*■ 




6 




k-l 




to 





ESPECIALLY MYCTOPHOIDS 



109 



epural is present anterior to the postero-dorsal extension of the stegural. The 
number of principal rays is 19 of which 17 are branched. There are at least 10 
accessory rays epaxially and 10 hypaxially. 

Vertebral column. The vertebral column is shown in the restoration, Text-figure 
47. In the specimens examined the vertebral number ranges from 35 to 40, and the 
centra are constricted. Of the total number of vertebrae approximately 18 are 
precaudal, the remainder (approximately 20) are caudal. The lateral face of each 
centrum bears a pair of longitudinal ridges between which is a rectangular depression. 
The neural arches occupy the entire dorsal surface of each centrum with the neural 
spines arising near to the posterior end of each arch. The suture between the neural 
arch and the corresponding centrum is clearly visible and the most anterior neural 
arches have a tendency to be separated from the centra during fossilization. Pre- 
zygapophyses and postzygapophyses are present on the neural arches but neither 
are particularly prominent. 



n.a.pu2 




Fig. 48. Eurypholis boissieri Pictet. Caudal fin skeleton in lateral view. 



The precaudal vertebrae bear paired ventro-lateral transverse processes on the 
anterior end of each centrum. Pleural ribs articulate with these parapophyses, and 
the ribs are moderately long and flattened. The ribs decrease in size posteriorly. 
All of the haemal arches, except for the first, are drawn out posteriorly into haemal 
spines. Epineurals and epipleurals are associated with the first 30 vertebrae. 

Scales. The body is devoid of scales except for a row along the course of the 
lateral line, and three large dermal scutes between the occiput and the dorsal fin in 
the mid-dorsal line of the body. These dorsal scutes are oval in shape and bear 



no UPPER CRETACEOUS TELEOSTS 

median longitudinal ridges on their dorsal surfaces. Each scute is ornamented 
dorsally with concentric ovals of small raised tubercles. The scutes overlap slightly 
and extend almost back to the origin of the dorsal fin. These scutes definitely give 
the appearance of having had a sensory canal associated with them, particularly as 
there is evidence of a median dorsal sensory canal which passed on to the supra- 
occipital and may possibly have passed through these three dorsal scutes. 

The scales marking the passage of the lateral line canal begin just posterior to the 
dorsal supracleithral region and continue back along the length of the body to 
terminate on the caudal peduncle. There are approximately the same number of 
scales in the row as there are vertebrae, i.e. 40. The scales are irregularly triangular, 
the base of each triangle being directed posteriorly. Each scale has a single deep 
indentation posteriorly, and here it overlaps the anterior region of the next scale in 
the row. On the anterior region of each scale there is a raised median ridge which is 
continued as a projection posteriorly above the anterior region of the hind indenta- 
tion. The outer surface of each scale is ornamented with small scattered tubercles 
except in the anterior region which is overlapped by the preceeding scale. The 
individual scales were perforated by the sensory canal which also ran within the 
posterior indentation of each scale. The projecting spine on each scale becomes 
more pronounced near the posterior end of the body, especially on the caudal 
peduncle where the terminal two or three scales bear prominent laterally projecting 
flanges. 

Eurypholis pulchellus (Woodward) 
(Text-figs. 49-54) 

1888a Enchodus lewesiensis (Mantell) ; Woodward : 315, pi. 1, fig. 6 (errore). 

1901 Enchodus pulchellus Woodward : 193, pi. 11, figs. 2, 3. 

1903 Enchodus pulchellus Woodward ; Woodward : 62, pi. 14, figs. 9-11, text-fig. 14. 

Diagnosis (emended). Eurypholis in which the maxilla does not appear to be 
toothed. A single major tooth row on the dentary with no indications of a marginal 
row. Mandible long and narrow with its greatest depth being about one-fifth of the 
total length. Preoperculum deep and narrow and forwardly inclined. Operculum 
slightly deeper than broad. 

Holotype. Specimen number P. 1703 in the B.M.N.H., an isolated head exposed 
from the left side, from the Turonian of south-east England. 

Material. The holotype and one other specimen from the B.M.N.H., number 
P. 10984, also from the Turonian of south-east England. 

Remarks. This species was originally placed by Woodward in the genus 
Enchodus. At first (1888a : 315, pi. 1, fig. 6) Woodward included it in the species 
Enchodus lewesiensis, later however he erected a new species for its reception, E. 
pulchellus Woodward (1901 : 193). The holotype figured by Woodward (1901, 
pi. 11, fig. 2 ; and 1903, pi. 14, fig. 9) is the remains of a head on a block of chalk 
with the left side exposed. The specimen was mechanically prepared from the right 
side, thus exposing much of the neurocranium and showing it to be unlike the neuro- 



ESPECIALLY MYCTOPHOIDS 




inf.sc 



epo 



med.s.c 



st.s.c 



10 mm 



Fig. 49. Eurypholis pulchellus (Woodward). Neurocranium in dorsal view. The broken 
line on the right side of the figure indicates the course of the sensory canals. From 
B.M.N.H. number P.1703. 



[12 UPPER CRETACEOUS TELEOSTS 

cranium of Enchodus lewesiensis. Apart from the size difference the cranial roof is 
practically identical to that of Eurypholis boissieri. In order to elucidate these 
points a second specimen was prepared in acid, P. 10984, from the Holaster sub- 
globosus zone of Hitchin, Hertfordshire. This, like the holotype, was exposed from 
the right side. The neurocranium which was recovered after acid treatment was 
incomplete only in respect of the roof region. Thus using the two specimens P. 1703 
and P. 10984 a complete description of the neurocranium and jaws was obtained. 
Although the body is missing from both specimens, the head supplies ample justifica- 
tion for transferring the species to the genus Eurypholis. 

Description. Neurocranium. The neurocranium is shown in dorsal, ventral, 
lateral and posterior views in Text-figures 49-52. The cranial roof is flattened and 
attains its widest point behind the orbits in the sphenotic region. The frontals cover 
practically the whole of the skull-roof and meet in the mid-dorsal line in a straight 
suture. The interorbital region of the skull-roof is slightly depressed and more 
sparsely ornamented. The frontal contacts the dorsal surface of the sphenotic and 
spreads partially above it. The entire roof is covered by an ornamentation of raised 
tubercles, and where the sensory canals passed within the bone the ornamentation is 
accentuated, the tubercles being raised on bony ridges. The ornamentation 
radiates outwards from the centre of ossification. Anteriorly the frontals taper from 
in front of the orbits and are inserted into the rear edge of the mesethmoid. Postero- 
laterally the frontal contacts the pterotic, and posteriorly the parietal and the 
supraoccipital. 

The supraoccipital is a small median bone contacting the frontals anteriorly and 
separating the parietals. In the mid-line of the posterior face of the supraoccipital 
is a small backwardly directed crest. 

The parietals are narrow, transversely elongated strips of bone. Anteriorly the 
parietals are bordered by the frontals and posteriorly by the epiotics. The tuber- 
cular ornamentation of the frontal bone is continued on to the parietal and the 
supraoccipital. A shallow groove present on each parietal bone is continuous 
with a shallow transverse groove on the supraoccipital. Behind the transverse 
groove on the parietal the hind border of the skull-roof is composed of a further strip 
of the parietal bounded posteriorly by the epiotic. This posterior part of the 
parietal is unornamented, probably indicating that the epaxial musculature en- 
croached on to the postero-dorsal surface of the skull-roof. 

The pterotic forms the postero-lateral border of the skull-roof, joining the sphenotic 
anteriorly, the frontal and parietal medially, and the epiotic posteriorly. The dorsal 
surface of the pterotic is heavily ornamented, the tubercles being joined by inter- 
vening ridges of bone to give an irregular latticed appearance. 

The sphenotic forms the postero-dorsal region of the orbit and contacts both the 
frontal and the pterotic. The dorso-medial region of the bone is ornamented 
similarly to the pterotic, but more laterally the surface becomes smoother. 

The course of the sensory canals on the skull-roof is indicated by tubes within the 
bone and open grooves on the surface. The supraorbital sensory canal passed for- 
wards within the frontal and emerged anteriorly on the lateral edge of the frontal. 



ESPECIALLY MYCTOPHOIDS 113 

Posteriorly this tube opens above the sphenotic into a shallow triangular depression 
devoid of ornamentation. The infraorbital sensory canal crossed the dorso-lateral 
surface of the sphenotic before it entered the bone. It opened again more medially 
into the triangular depression already noticed in association with the supraorbital 
sensory canal. The otic branch of the infraorbital sensory canal entered the pterotic 
posteriorly and passed forwards within the bone also to open anteriorly into the 
triangular depression. A postero-lateral pore is visible in the pterotic and marked 
the passage of the preopercular sensory canal. A supratemporal sensory canal 
appears to have been present and may have connected with the otic branch of the 
infraorbital sensory canal within the posterior region of the pterotic. The sensory 
canal passed medially into the parietal where it opened into a groove on the surface. 
This groove is continuous across the supraoccipital, and it is assumed that the supra- 
temporal sensory canal was continued within this groove. 

The mesethmoid is a short broad bone produced laterally into wings which are 
attached to the dorsal surface of the vomer. The lateral edges of these wings serve 
as the articulatory points for the anterior regions of the palatine bones. The 
anterior edge of the mesethmoid is transversely arranged and the premaxillary 
pedicels articulated with it. 

The lateral ethmoids are small but stout bones, especially ventrally where they 
provide the posterior articulatory point for the palatine. Laterally the lateral 
ethmoid provides a further articulatory point for the lachrymal, and dorsally it is 
attached to the underside of the frontal. 

The vomer is separated from the mesethmoid postero-medially but attached to it 
anteriorly and laterally. The vomer has corresponding lateral wings to those seen 
on the mesethmoid. A short vomerine process extends back along the ventral 
surface of the parasphenoid. 

The parasphenoid is curved below the orbit and is little expanded ; the lateral 
regions being simply flattened and inclined ventrally producing a median, ventral 
longitudinal groove. The ascending process of the parasphenoid curves dorsally to 
meet the prootic, and has a thickened anterior edge. The foramen which trans- 
mitted the internal carotid artery into the base of the orbit lies near to the antero- 
ventral edge of the ascending process. The posterior part of the parasphenoid, 
below the myodome, bears a large mid-ventral, curved keel. The parasphenoid 
joins the ventro-lateral regions of the basioccipital and forms the floor and walls of 
the myodome. The myodome has a tiny posterior opening between two lateral 
sheets of parasphenoid below the occipital condyle and above the ventral paras- 
phenoid keel. 

The occipital condyle is circular and concave and composed of basioccipital 
ventrally and of the exoccipitals dorsally. The basioccipital contacts the exoccipitals 
dorsally, the parasphenoid ventrally and the prootics anteriorly. On the postero- 
lateral region of the bone near to the condyle is a group of pits and ridges. Internally 
the basioccipital forms the walls of the posterior part of the otolith chamber, and its 
ventral surface provides part of the roof of the myodome. Antero-dorsally the 



II 4 



UPPER CRETACEOUS TELEOSTS 




spo 



epo 



10 mm 



soc 



exo 



Fig. 50. Eurypholis pulchellus (Woodward). Neurocranium in ventral view. 



ESPECIALLY MYCTOPHOIDS 



115 



basioccipital contacts the prootic bridge above the myodome. The antero-dorsal 
extent of the bone provides a partial separation of the two otolith chambers. The 
otolith chambers themselves extend anteriorly to end in a recess in the prootic just 
posterior to the internal opening of the facial foramen. 

The prootics divide to produce an internal lamina which meets its partner in the 
mid-line forming the prootic bridge, and an outer lamina which curves laterally in 
the dorsal region and provides part of the wall of the cranial cavity. Posterior to the 
prootic bridge the otolith chambers communicate with the cranial cavity through a 
large oval fenestra. On the orbital face, the prootic meets the sphenotic and 
pleurosphenoid dorsally. The prootic contains the trigemino-facialis chamber 
which is divided by bone into a medial pars ganglionaris and a lateral pars jugularis. 
The pars jugularis is a very short horizontal canal on the lateral face of the prootic 
which opens anteriorly in the hind wall of the orbit. There is a single foramen lead- 
ing from the pars ganglionaris into the pars jugularis, the facial foramen, which 
transmitted the hyomandibular trunk of the facial nerve. The palatine branch of 
the facial nerve passed through a foramen medial to the facial foramen down into the 
myodome. The hyomandibular nerve passed out through the posterior opening of 
the pars jugularis, which also transmitted the jugular vein and the orbital artery. 
The artery and vein were the only structures which passed through the anterior 
opening of the pars jugularis. The trigeminal and remaining components of the 
facial nerve passed through two foramina in the orbital face of the prootic, dorso- 
medial to the anterior opening of the pars jugularis. The larger of the two foramina 
is the trigeminal foramen and presumably transmitted all of the remaining branches 
except the superficial ophthalmics. These latter components passed out through a 
small foramen dorso-medial to the trigeminal foramen. 



d.f spo 




Fig. 51. Eurypholis pulchellus (Woodward). Neurocranium in lateral view. The 
inset shows the prootic in antero-lateral view. From B.M.N. H. number P. 10984. 



I 10 



ri'l'ER CRETACEOUS TELEOSTS 



The exoccipital bones meet in the mid-line of the posterior face of the neuro- 
cranium below the foramen magnum and just contact each other above it, com- 
pletely enclosing the foramen magnum. Ventrally each exoccipital contributes 
towards the composition of the occipital condyle, and laterally forms part of the 
cranial wall, contacting the intercalar, the pterotic and the prootic. The foramen 
through which the vagus nerve passed is a prominent hole on the lateral face of the 
neurocranium in front of the condyle. Between the vagus foramen and the occipital 
condyle two or three small foramina transmitted the occipital nerves to the anterior 
region of the vertebral column. Anterior to the vagus foramen on the lateral wall of 
the cranial cavity a further small foramen is present, which transmitted the glosso- 
pharyngeal nerve. Dorsally the exoccipital takes part in the formation of a large 
concave depression together with the prootic and the pterotic ; this depression 
housed musculature associated with the branchial apparatus. On the posterior face 
of the neurocranium thickenings of the exoccipital pass laterally to the intercalar 
and dorso-laterally towards the epiotic. 

The pterotic forms almost the entire hyomandibular facet which is an elongate 
groove on the ventro-lateral surface of the bone. The facet itself is horizontal with 
the posterior region rising very slightly. Above the hyomandibular facet the 
dilatator fossa is visible in lateral aspect, its posterior border formed by the pterotic. 
On the posterior face of the neurocranium a post-temporal fossa is also present 
and the pterotic forms the lateral parts of the roof, wall and floor of this fossa. 
The remainder of the roof is composed medially of frontal and parietal with a 
minute portion of epiotic at the extreme posterior end. The post-temporal fossa 
is not particularly prominent and is only visible in posterior view. 

The sphenotic forms the postero-dorsal orbital region, the most anterior region of 
the elongated hyomandibular facet and the anterior border of the dilatator fossa. 
The dorsal part of the sphenotic projects laterally and from the underside of this 
projection a strut extends ventro-medially to contact a corresponding strut passing 
dorso-laterally from the prootic, the dorso-lateral strut of the prootic being a direct 




Fig. 52. Eurypholis pulchellns (Woodward). Neurocranium in posterior view. 
From B.M.N.H. number P.10984. 



ESPECIALLY M YCTOPHOIDS 117 

continuation of the slip of bone forming the lateral wall of the pars jugularis. The 
anterior region of the hyomandibular facet is immediately behind the strut on the 
sphenotic. 

The epiotic forms much of the posterior face of the neurocranium, and extends 
on to the extreme posterior end of the skull-roof. Laterally the epiotic enters 
into the composition of the post-temporal fossa forming the medial parts of the 
posterior end, i.e. floor, wall and roof. In the floor of the post-temporal fossa it 
contacts the pterotic and on the posterior face of the skull, below the opening of the 
fossa, it meets the intercalar. 

The intercalar is a moderately large bone forming part of the cranial wall itself. 
It is bent mesially into a right angle and forms the dorso-lateral continuation of the 
strut produced on the exoccipital laterally. On the posterior face of the skull it 
contacts the exoccipital, the epiotic, and dorso-laterally, below the post-temporal 
fossa, it joins the pterotic. On the lateral face of the neurocranium it contacts the 
exoccipital and the pterotic below the hind end of the hyomandibular facet. A 
posterior projecting knob of bone is present on the intercalar representing the point 
at which the ventral limb of the post-temporal bone attached to the skull. 

The pleurosphenoids lie in the dorsal region of the posterior face of the orbit, and 
contact the frontals above, the sphenotics laterally and the prootics ventrally. The 
superficial ophthalmic nerves crossed the face of the pleurosphenoid in a slight 
groove near the medial edge of the bone. 

Infraorbital bones. The infraorbital bones are shown in lateral view in Text-figure 
54 and differ in no way from those of the type species, Eurypholis boissieri (p. 102). 

Hyopalatine bones. The hyopalatine bones are shown in medial view in Text- 
figure 53. The hyomandibular is broad, curved and inclined forwards ventrally. 
The head is elongated antero-posteriorly. The opercular process on the hind edge of 
the bone is prominent and occurs just within the lower half of the bone. A promi- 
nent crest on the lateral face arises near the anterior region of the head and curves 
postero-ventrally towards the posterior edge, terminating near the ventral extremity. 
The hyomandibular tapers ventrally. Anteriorly the hyomandibular is thin, 
expanded and overlain laterally by the metapterygoid. The hyomandibular branch 
of the facial nerve entered through a large antero-dorsal foramen on the medial 
face of the bone. Within the bone the nerve divided into the opercular branch, 
which emerged through a foramen immediately above the opercular process, and 
the mandibular branch which passed through a foramen situated below the opercular 
process. The mandibular nerve continued ventrally in a shallow groove on the 
antero-medial edge of the preoperculum on to the mandible ventrally, below the 
articular facet. 

The symplectic is short, stout and inclined forwards ventrally. Its ventral region 
lies in a wide groove on the postero-medial part of the quadrate and dorsally it ends 
close to the ventral extremity of the hyomandibular, to which, in life, it was 
connected by cartilage. 

The quadrate is large and triangular, with a prominent transversely arranged 
condyle. The bone thins dorsally, but its anterior and posterior regions are thick 



n8 



UPPER CRETACEOUS TELEOSTS 



and well ossified. The medial groove containing the symplectic is broad and deep 
and ends just postero-dorsal to the condylar region. The posterior edge of the 
quadrate lies in a groove on the antero-ventral edge of the preoperculum and is over- 
lapped laterally and medially by the preoperculum. Anteriorly the quadrate is 
rigidly attached to the postero-medial face of the ectopterygoid. 

The ectopterygoid is in two portions joined ventrally, giving a V-shaped dorsal 
trough between the two components. The ventral edge of the bone is curved and 
bears teeth. The more posterior teeth are not strictly on the ventral edge of the 
bone but their bases have moved upwards on to the lateral face of the ectopterygoid. 
The teeth, however, form a continuous series and are acute, laterally compressed, 
and hollow. The tooth size decreases towards the posterior end of the bone. All of 
the teeth have the same angular relation to the supporting bone, i.e. at 90° to it. 
Therefore the more posterior teeth are directed forwards due to the postero-ventral 
curvature of the ectopterygoid. All of the teeth are marked by fine longitudinal 
striations and the bases are expanded and ornamented with small pits and fine striae. 
The ectopterygoid contacts the postero-ventral surface of the palatine and extends 
along the medial surface of the latter bone. The palatine thus sits in the groove on 
the dorsal surface of the ectopterygoid. The medial flange of the ectopterygoid is 



hmh 



f mes vo 



pall 



op.p hm 




sop 



g.md.VII den 



Fig. 53. Eurypholis pulchellus (Woodward). Hyopalatine and opercular bones and 
mandible of the left side in medial view. From B.M.N.H. number P. 10984. 



ESPECIALLY MYCTOPHOIDS 



119 



larger than the lateral flange and unites dorsally with the lateral edge of the endo- 
pterygoid. Posteriorly it is the medial flange which articulates with the quadrate, 
the lateral flange decreasing in size and merging with the medial one posterior to the 
last tooth. 

The metapterygoid is thin and membranous and missing in parts. Dorsally it lies 
against the lateral face of the hyomandibular and its outer face bears a shallow 
obliquely inclined crest which forms the anterior edge of the bone near the ventral 
limit of the metapterygoid. Antero-ventrally the metapterygoid is attached to the 
ectopterygoid and the endopterygoid, and ventrally to the dorsal edge of the 
quadrate. 

The endopterygoid is a narrow sheet of bone attached to the innermost edge of the 
ectopterygoid and lying in the roof of the mouth. The dorsal surface is concave 
below the orbit, but its convex ventral surface is smooth and untoothed. 

The palatine is a prominent, bulbous structure with its posterior end resting in the 
trough on the dorsal ectopterygoid surface. The palatine bears one enlarged anterior 
tooth set at 90 ° to the axis of the bone. The tooth base merges imperceptibly with 
the palatine bone and the tooth is identical in form to those on the ectopterygoid. 
Posteriorly the palatine is extended dorsally forming a facet for the reception of 
the ventral region of the lateral ethmoid. Dorso-medially the palatine has an 
elongated oval facet which houses the lateral wing formed from the fused lateral 
extensions of the mesethmoid and vomer. 




pop.s 



Fig. 54. Eurypholis pulchellus (Woodward). Skull of the holotype (B.M.N.H. 
number P. 1703) in lateral view. 



r2o UPPER CRETACEOUS TELEOSTS 

Dermal upper jaw. The dermal upper jaw is shown in lateral view in Text-figure 
54. The premaxilla is large and has a prominent pedicel anteriorly, lying at 90 ° to the 
lateral region of the bone. The premaxillary pedicels meet in the mid-line and lie in 
the same plane as the anterior part of the skull-roof and form a direct continuation 
of it, the pedicels articulating with the anterior edge of the mesethmoid. Each pedicel 
contains an oval fenestra through which the enlarged anteriormost tooth of the dent- 
ary projects when the jaws are closed. Between the premaxillary pedicel and the 
lateral face of the bone there is a groove running to the anterior edge of the snout, 
this groove is continuous with the anterior opening of the supraorbital sensory canal 
on the lateral edge of the frontal. The premaxilla is closely associated with the 
lateral face of the palatine and extends back to a point below the centre of the orbit. 
The lateral face of the premaxilla is triangular and ornamented with four or five rows 
of bony tubercles which radiate backwards and outwards from the anterior snout 
region. Postero-dorsally the lateral face is overlapped by the ventral expansion 
of the lachrymal bone, and a lateral groove devoid of ornamentation extends 
anteriorly. This groove may have contained an anterior prolongation of the 
infraorbital sensory canal. The premaxilla decreases in depth below the lachrymal 
and tapers posteriorly. 

The maxilla enters the gape behind the premaxilla and along its whole length is 
closely attached to the premaxilla making the demarcation line between the two 
bones difficult to define. The maxilla is continued on the internal face of the pre- 
maxilla and is firmly fused to it. The maxilla is smooth and rod-like with no 
ornamentation. There is a single marginal row of teeth on the premaxilla which does 
not extend on to the maxilla. The teeth are small, evenly spaced, acutely pointed 
and merely seem to be a slightly enlarged marginal row of bony tubercles similar 
to the tubercular ornamentation on the other dermal bones of the skull. 

Mandible. The mandible is shown in medial and lateral views in Text-figures 53, 
54. The dentary forms almost the whole of both dorsal and ventral margins of the 
mandible and over half of the lateral face. The posterior edge of the dentary is 
deeply indented into a V-shape. The mandible is elongate and relatively shallow 
with the ventral region strongly inflected towards the mid-line. Dorsally the dentary 
bears a medial flange on which the teeth are borne in a single row. The teeth are like 
those found on the palato-pterygoid arcade, i.e., conical, acutely pointed, hollow, 
slightly recurved, laterally compressed, and with expanded bases. The dentary also 
bears a ventro-medial flange which diverges from the dorsal flange in the anterior 
mandibular region. The deep V-shaped groove left between the dorsal and ventral 
flanges on the internal face of the mandible served to house the remains of Meckel's 
cartilage. 

The articular facet is concave and transverse, limited medially by a small vertical 
flange, and laterally by a large vertical upgrowth of the articular. This lateral 
flange completely covers the articular facet so that it is not visible in lateral view. 
The articular extends anteriorly to occlude the posterior indentation of the dentary, 
and is inserted on to the medial face of the dentary. The mandibular sensory canal 
entered the articular region of the mandible behind the articular facet and passed 



ESPECIALLY MYCTOPHOIDS 121 

within the bone ventral to the facet. The tube which contained the sensory canal 
passes forwards on the ventro-lateral region of the articular and opens into a shallow 
groove on the posterior region of the dentary. The lateral face of the mandible is 
ornamented with bony tubercles which run in regular lines radiating backwards from 
the symphysial region on the dentary, and forwards from the facet region on the 
articular. The ornamentation is more pronounced antero-ventrally where the 
tubercles are raised on bony ridges. 

The angular is a small knob of bone forming the postero-ventral angle of the mand- 
ible. 

Opercular bones. The opercular bones are shown in medial and lateral views in 
Text-figures 53, 54. The preoperculum is deep, narrow and forwardly inclined 
ventrally. Dorsally it ends behind the head of the hyomandibular where it rests 
against the posterior face of the crest on the hind edge of the hyomandibular. More 
ventrally the anterior edge of the preoperculum is excavated vertically, so that a 
slight concavity is produced into which the posterior edge of the quadrate fitted. 
Ventrally again the preoperculum is produced into a small backwardly projecting 
spine. The preopercular sensory canal ran within a tube in the bone opening dors- 
ally by a single pore, and ventrally by two or three pores. Two of the ventral pores 
open posteriorly on the region of the spine, whilst the most anterior pore opens 
anteriorly on to the articular. The lateral face of the preoperculum is ornamented 
with characteristic bony tubercles which are more prominent in the region above the 
sensory canal and along the posterior margin of the bone. Ventrally on the pre- 
opercular spine the tubercles are raised on ridges of bone. 

The operculum has a straight anterior edge covered by the preoperculum laterally. 
The dorsal border of the operculum is rounded but ventrally the edge is concave. 
From the articulatory facet midway along its anterior edge an internal horizontal 
strengthening ridge crosses the bone and terminates on the posterior angle where it is 
produced into a short spine. Laterally the operculum is ornamented with radiating 
rows of bony tubercles arising from the point of articulation with the hyomandibular. 

The suboperculum is large with its dorsal part lying medial to the ventral region of 
the operculum. The lateral face of the bone is ornamented with a regular arrange- 
ment of bony tubercles. 



Genus SAURORHAMPHUS Heckel, 1850 

Diagnosis (emended). Eurypholidae in which the head is elongated, especially in 
the preorbital region where it is extended into a prominent rostrum. Vertebral 
number 42 of which 22 are caudal. Body very shallow, never exceeding the depth of 
the head, and extended by an increase in the length of the individual centra especially 
precaudally. Pelvic fins and pectoral fins of equal size ; pelvic fin almost sub- 
thoracic in position. Mid-dorsal scutes 6 or 7 in number, overlapping each other 
in advance of the dorsal fin. 

Type species. Saurorhamphus freyeri Heckel. 



i22 UPPER CRETACEOUS TELEOSTS 

Remarks. The genus Saurorhamphus is known by this single species originally 
described by Heckel (1850). Woodward (1901 : 210) transferred it to the genus 
Eurypholis but this was later shown by d'Erasmo (1912 : 87) to be incorrect, Sauror- 
hamphus representing a distinct genus closely associated with Eurypholis. This 
latter conclusion is substantiated in the following description. 



Saurorhamphus freyeri Heckel 

(Text-fig. 55) 

1850 Saurorhamphus freyeri Heckel : 217, pis. 18, 19. 
1946 Saurorhamphus freyeri Heckel ; d'Erasmo : 92, text-fig. 24. 
d'Erasmo lists the earlier references. 

Diagnosis (emended). Saurorhamphus of standard length up to 18 cm. Length 
of head with opercular apparatus just over three times the depth of the head, equal 
to one-third of the standard length. Maximum depth of the mandible equal to 
one-seventh of the total length. Maxilla forms one-third of the oral border of the 
mouth and is untoothed. Operculum broader than deep. Dorsal fin in the mid- 
trunk region with 14 rays. Anal fin remote, with 15-16 rays arising well behind the 
dorsal fin, close to the caudal fin. 

Holotype. Complete flattened fish wanting snout, in the Natural History section, 
Prirodoslovni Muzej Slovenije, Ljubljana, Jugoslavia, from the Lower Cenomanian of 
Comen, near Trieste. 

Material. Only the holotype in Ljubljana was examined since none of the speci- 
mens used by d'Erasmo (1912, 1946) could be traced. 

Description. Neurocranium. The posterior region of the skull-roof is the only 
part of the neurocranium which is at all well shown. The frontals occupy practically 
the entire roof, reaching back almost to the occiput. They are long and narrow and 
attain their greatest width at the rear end of the orbit above the sphenotic. The 
skull-roof is flat and ornamented with a regular arrangement of bony tubercles which 
radiate in all directions from the centre of ossification on the frontal. Anteriorly the 
frontal tapers gradually from the front end of the orbit and is inserted into the hind 
end of the mesethmoid. 

The supraoccipital is small and separates the transversely arranged parietals. It 
bears no indication of having had a supraoccipital crest, but each parietal bears a 
shallow, transverse, ornamented crest. The tubercular ornamentation seen on the 
frontals is continued on to the parietals and the supraoccipital. 

The pterotic forms the postero-lateral border of the skull-roof and dorsally 
appears as a narrow elongated strip of bone extending anteriorly on to the dorsal 
surface of the sphenotic. Posteriorly it projects slightly behind the level of the 
occiput and its dorsal surface is ornamented with bony tubercles. 

The sphenotic forms the postero-dorsal region of the orbit and projects laterally 
from beneath the frontal. Unlike the other roofing bones the sphenotic is unorna- 
mented. 



ESPECIALLY M YCTOPHOIDS 123 

The epiotics are present on the occipital border of the skull-roof as smooth hemi- 
spherical protuberances providing an articulatory surface for the dorsal limb of the 
post-temporal. 

Infraorbital bones. Traces of a circumorbital series of bones can be seen. The 
bones which are visible bear an external ornamentation of bony tubercles. Beneath 
the orbit there is a long shallow bone probably representing the first infraorbital, and 
behind this there is a larger bone (2nd infraorbital) which is angled and slightly more 
expanded and which shows the tube which housed the sensory canal. 

Jaws and suspensorium. The hyomandibular is large with a greatly elongated 
head. Ventrally the hyomandibular curves forwards so that the mandibular 
articulation lies below the centre of the bone. The lateral face of the hyomand- 
ibular is covered to a large extent by the postero-dorsal region of the metapterygoid. 

The anterior region of the head is only present as an impression in the matrix so 
that description is rendered difficult. The ectopterygoid would seem to be greatly 
extended as is also the palatine. The ectopterygoid bears a longitudinal series of 
evenly spaced teeth, but the dentition of the palatine is difficult to determine. At 
the anterior end of the rostrum an impression of a palatine tooth is visible. It is large 
and extends downwards over the impression of the lower jaw. Whether further 
palatine teeth were present cannot be observed on the specimen. No corresponding 
enlarged tooth of the dentary at the symphysis is present. Within the orbit, below 
what appears to be parasphenoid, the endopterygoid forms a flattened lamina of 
bone. The parasphenoid in the orbital region is slightly expanded ventro-laterally. 

The dermal upper jaw is only present as an impression in the rock. The pre- 
maxilla appears to be elongated and large, extending posteriorly to the orbit and 
bearing a lateral ornamentation of bony tubercles. The maxilla enters the oral edge 
of the upper jaw below the orbit. The premaxilla is toothed marginally but the 
maxilla appears to be untoothed. 

The mandible is prominent, elongate, relatively shallow and with a constricted 
symphysis. The lateral face is covered with a dense ornamentation of raised 
bony tubercles. The suture between the dentary and the articular could not be 
observed in the specimen. A dorsal extension of the articular lateral to the articular 
facet extends upwards and hides the facet and the quadrate condyle in lateral view. 
The retroarticular process is insignificant, and the coronoid process is low. The 
mandibular dentiton was not observed in the specimen. 

The preoperculum is deep and narrow although there are anterior and posterior 
expansions ventrally. The anterior edge of the bone follows the anterior curvature 
of the hyomandibular and the quadrate, and abuts against the posterior edges 
of both of these bones. Post ero- ventrally there is a slight backwardly projecting 
spine. The lateral face of the preoperculum is ornamented with a uniform covering 
of bony tubercles. The operculum is large and about as deep as long with a straight 
anterior edge. The facet for the articulation with the hyomandibular is mid- 
way down the anterior edge of the bone. The dorsal edge of the operculum is 
rounded and the margin is produced posteriorly into a short spine. Ventrally the 



124 



UPPER CRETACEOUS TELEOSTS 



c 









+J 




flj 








4J 




^d 




CO 


d 


0> 


o 




'-P 


o 


o 




Cm 


o 


- — 1 


l-H 

o 

-4-J 


CO 
CD 

3 


CO 


O 


4) 


[/} 


ft 


ri 




01 




IH 


^ 


o 


CD 


T3 


O 

X 


3 




T3 




CD 


h 


+J 


<& 

*, 


"a 

o 




CO 

CD 


►ss 




f 


o 

CO 


« 




-si 




~ 




o 




^ 




S 




<3 




■'. 





fa 



ESPECIALLY MYCTOPHOIDS 125 

margin of the operculum is straight and slightly oblique. An inner strengthening 
ridge runs from the region of the facet anteriorly and traverses the medial face of 
the bone to terminate on the small posterior spine. The suboperculum is prominent 
and like the operculum bears an inner strengthening ridge. Both bones are orna- 
mented with tubercles, those on the operculum radiating outwards from the anterior 
edge of the bone in the region of the opercular facet. 

Pectoral and pelvic girdles and fins. The post-temporal has an oval upper limb whose 
dorsal surface is ornamented with tubercles. The supracleithrum could not be seen 
since it lies medial to the upper opercular region. The cleithrum is large and 
ventrally it is expanded both anteriorly and posteriorly. The anterior end of the 
cleithrum almost reaches the preoperculum, while the posterior end projects back- 
wards beyond the insertion of the pectoral fin. Ventrally the lower margin of the 
cleithrum is inflected towards the mid-line. The lateral face of the cleithrum bears 
the same tubercular ornamentation as the opercular bones and the tubercles are 
more pronounced ventrally. The pectoral fin arises just below the middle of the 
flank and appears to consist of approximately 12 rays forming a moderately large fin. 

The pelvic bones are prominent and each consists of a thin sheet of bone roughly 
triangular in shape. The pelvic bones lie immediately posterior to the cleithrum and 
thus can be said to be sub-thoracic in position. The pelvic fins are as large as the 
pectoral fins and are composed of 7 or 8 stout rays. 

Median fins and tail. The dorsal fin is situated behind the mid-point of the body 
and is composed of 14 rays. The first ray is small and unbranched, the second ray is 
the longest and the remaining rays decrease in length. All the rays are segmented, 
and except for the first are branched. The pterygiophore number parallels the 
number of fin rays, and the first 3 proximal radials are expanded, especially the 
first which bears an expanded anterior median keel. The anal fin is remote and 
consists of 15 or 16 rays ; unlike the dorsal fin, none of the proximal radials is 
expanded. 

The caudal fin consists of 19 principal rays. Accessory caudal rays are present in 
front of the principal rays both above and below the fin, there being between 18-20 in 
all. 

Vertebral column. The vertebral column has some 41 or 42 vertebrae of which 
approximately 22 are caudal. In the anterior precaudal region the first few centra 
are long and narrow and mesially constricted. The centra begin to shorten in length 
at about the 10th precaudal vertebra and by the 14th precaudal vertebra they have 
become reduced to a length which remains constant to about the 13th caudal (the 
34th vertebra). From this vertebra to the end of the column the centra become 
further reduced in length and depth so that at the caudal peduncle the centra are very 
small and practically square. The neural arches are prominent and the line of 
articulation between the arch and the corresponding centrum is clearly visible. The 
neural spines are short and arise from the posterior region of each neural arch. In 
the precaudal region the neural spines are at their longest and exhibit a slight 
backward curvature. Below the dorsal fin the spines shorten and straighten, and 



126 UPPER CRETACEOUS TELEOSTS 

along the caudal region become progressively shorter. Small transverse processes 
are present on the precaudal vertebrae, supporting fine pleural ribs. The haemal 
spines are short but slightly more curved than the corresponding neural spines of the 
caudal region. 

Intermuscular bones are present along practically the entire length of the body, 
epineurals dorsally and epipleurals ventrally. 

Scales. Scales are only present along the lateral line and in the mid-dorsal line 
behind the occiput. These latter scutes may be as many as six in number and are 
large, ovoid, and ornamented with a median longitudinal ridge as well as concentric 
rings of minute tubercles. These scutes may have contained a branch of the sensory 
canal system from the head. The third scute in particular shows a postero-dorsal 
pore in the median ridge. The lateral line scales begin just posterior to the post- 
temporal bone where they are very large, more posteriorly they decrease in size 
progressively. The scales overlap and there are approximately 42 scales in the row. 
Each scale is irregularly triangular with the base of the triangle directed posteriorly. 
This posterior edge bears a deep, narrow indentation. This indentation occurs 
where the scale overlaps the anterior region of the succeeding scale. The lateral 
surface of each of the scales is ornamented with raised tubercles except in the over- 
lapped portion. 

Suborder HALECOIDEI 

Diagnosis. Head and body shallow to slightly deepened. Parietals separated in 
the mid-line by the supraoccipital. Post-temporal fossa unroofed. Orbitosphenoid 
and basisphenoid absent ; supraorbital often present. Premaxilla toothed, with no 
ascending process, maxilla always with forwardly inclined teeth ; one supramaxilla 
present. Teeth on palatines, ectopterygoids and endopterygoids but not on vomer. 
Numerous branchiostegal rays. Mesocoracoid arch present ; pectorals inserted 
low on body. Pelvics abdominal or sub-thoracic. Vertebral elements incompletely 
fused ; epineurals numerous. Fins without spines. Caudal with 17 branched rays. 
Preural vertebra 1 fused with ural vertebra 1 ; ural vertebra 2 present as a terminal 
half-centrum. Scales present or absent, when present small and closely packed. 

Family HALECIDAE Agassiz, 1834 

Diagnosis. Head and body tending to become shortened and deepened. 
Maxilla forms at least half of the gape and bears large, forwardly inclined teeth ; 
characteristically small supramaxilla always present. Preopercular spine arising 
from the postero-ventral region of the preoperculum. Dorsal fin confined to the 
anterior half of the back. Anal fin always remote in position. 

Genus HALEC Agassiz, 1834 

Diagnosis (emended). Halecidae in which the head is short and undeepened and 
the body is slightly elongated. Supraorbital present. Jaw suspension vertical. 



ESPECIALLY MYCTOPHOIDS 127 

Branchiostegal rays at least 15 in number. Vertebrae between 40 and 46. Pelvic 
fins abdominal, equal in size to the pectorals. Epineurals on the first 30 vertebrae. 
Complete covering of minute scales, enlarged along the lateral line. 

Type species. Halec sternbergi Agassiz. 

Remarks. The family Halecidae was originally erected, as the name implies, to 
include the ' herring-like ' forms, the clupeoids and the salmonoids. The grouping 
together with the familial name was dropped later on, although used by Pictet 
(1850) and Davis (1887). The family is re-erected but only to contain three genera. 
Of the genus Halec, two species are considered, Halec eupterygius, preserved ' in 
the round ', and Halec haueri which is always laterally flattened. 

Halec eupterygius (Dixon) 
(Text-figs. 56-62) 

1837 Osmeroides lewesiensis Agassiz, 5 ; 1 : pi. 60b, figs. 3, 4 (errore). 
1850 Pomognathus eupterygius Dixon : 367, pi. 35, figs. 6, 7. 
1888a Pomognathus eupterygius Dixon ; Woodward : 318. 

1901 Halec eupterygius (Dixon) ; Woodward : 213. 

1902 Halec eupterygius (Dixon) ; Woodward : 50, pi. 13, text-fig. 11. 

Diagnosis (emended). Halec species reaching approximately 31 cm. in standard 
length. Length of the head with the opercular apparatus exceeds the depth of the 
trunk, and is contained four times in the length from the pectoral arch to the base of 
the caudal fin. Vertebrae 45 in number, at least 20 and possibly 22 being caudal. 
Dorsal fin triangular, with 12 rays, first rays stout and ornamented ; anal fin small 
and remote. 

Holotype. Imperfect fish in the Brighton Museum from the Turonian of south- 
east England. 

Material. Specimens in the B.M.N.H., listed by Woodward (1901 : 213-218), 
totalling some 30 specimens. Neurocranial description obtained from specimen 
numbers P. 36237, P. 4289, P. 10920, and 43392 all of which were prepared in acetic 
acid. The specimens are from the Turonian and Senonian stages of south-east 
England. 

Description. Neurocranium. The neurocranium is shown in dorsal, ventral, 
lateral and posterior views in Text-figures 56-59. The cranial roof is flattened and 
widest at the hind border of the orbit between the sphenotics. The orbital region is 
large and the parasphenoid is slightly curved below it. The major portion of the 
roof is formed from the frontals which extend back almost to the occipital border, and 
meet in the mid-line in a practically straight suture. Postero-laterally a small V- 
shaped indentation of the frontal marks the anterior limit of the post-temporal fossa. 
The frontal does not extend into the floor of the post-temporal fossa. Laterally the 
frontal overlaps the dorso-medial surface of the sphenotic, but the major part of the 
sphenotic still projects laterally from beneath it. The dorsal surface of the frontal 
is ornamented with finely tuberculated ridges radiating from the centre of ossifica- 
tion. This ornamentation is absent from the anterior and medial regions of the 



128 



UPPER CRETACEOUS TELEOSTS 



frontals. Anteriorly the frontal tapers very slightly to end at the rear of the short, 
broad, smooth mesethmoid. 

The supraorbital sensory canal ran within the frontal above the orbit, and gave off 
a branch which opened through a medial pore at the level of the centre of the 
orbit. The main supraorbital canal opened anteriorly on the edge of the frontal. 



so s c 




inf. s.c 



f.d.pt 



10mm 



Fig. 56. Halec eupterygius (Dixon). Neurocranium in dorsal view. The broken line 
on the left side of the figure indicates the course of the sensory canals. 



ESPECIALLY MYCTOPHOIDS 



129 



The supraoccipital is a small bone just appearing on the dorsal surface of the neuro- 
cranium. It is overlapped anteriorly by the frontals, and separates both the 
parietals and the epiotics. A small backwardly directed crest is produced which 
does not extend above the plane of the skull-roof. 



f.i.ca 




Qt.f.C 





1 


r 




fe.p.a 




^/spo 




^^■pro 






^f.hm 



f.o.n 



Fig. 57. Halec eupterygius (Dixon). Neurocranium in ventral view. 



'30 



UPPER CRETACEOUS TELEOSTS 



The parietals are narrow, transversely orientated strips of bone, at the hind end 
of the frontals. Each parietal bears a small transversely arranged parietal crest, 
but the remainder of the surface is smooth. 

The mesethmoid is short, broad and poorly ossified, but internally thin mem- 
branous sheets of bone extend back to the lateral ethmoids. Laterally the meseth- 
moid is attached to the vomer. 

The lateral ethmoids are flimsy bones representing merely superficial ossifications 
of cartilaginous components. The lateral and ventral regions of the lateral ethmoid 
are however more strongly ossified than the remainder. Ventrally the lateral 
ethmoids are attached to the postero-dorsal region of the palatine. 

A small supraorbital bone lies along the lateral edge of the frontal above the lateral 
ethmoid. The supraorbital is oval and finely ornamented with small tubercles. 

The vomer, which attaches to the undersurface of the mesethmoid, is a small 
shallow bone with a short posterior process. This posterior process extends back- 
wards below the anterior end of the parasphenoid. The vomer is expanded laterally 
and between the mesethmoid and vomer the head of the maxilla articulates. 

The parasphenoid extends for practically the entire length of the neurocranium. 
Below the orbit it widens considerably and consists of a pair of large ventro-laterally 
inclined flanges associated with the endopterygoids. The ascending processes of 
the parasphenoid curve upwards at the posterior end of the orbit to meet the 
prootics in the walls of the myodome. The anterior edge of each ascending process 
is thickened. The course of the efferent pseudobranchial artery is represented by 
a notch in the antero-ventral edge of the ascending process. The foramen for the 
internal carotid artery lies just behind the efferent pseudobranchial notch. Poster- 
iorly the parasphenoid joins the basioccipital. The parasphenoid terminates just 
in front of the hind end of the basioccipital and forms the complete floor of the 
myodome. 



f.hm 




Fig. 58. Halec enpterygins (Dixon). Neurocranium in lateral view. 



ESPECIALLY MYCTOPHOIDS 131 

The occipital condyle is circular and concave and formed of both basioccipital and 
exoccipitals. The basioccipital contacts the exoccipitals dorsally and the prootics 
antero-dorsally. Internally the basioccipital forms the walls of the posterior parts 
of the otolith chambers and the roof to the posterior myodome. The otolith 
chambers extend anteriorly to terminate in a recess within the prootic, lateral to the 
prootic bridge. 

The prootic extends medially to meet its partner forming the prootic bridge above 
the myodome. Dorso-laterally the prootic contacts the sphenotic within the 
anterior cup-shaped part of the hyomandibular facet. The orbital face of the 
prootic, above the prootic bridge, meets the sphenotic dorso-laterally, the frontal 
dorsally and the pleurosphenoid dorso-medially. The pars ganglionaris is a shallow 
medial recess in the prootic, lateral to the prootic bridge. The pars jugularis is a 
short horizontal canal lateral to the pars ganglionaris. Two foramina pass from the 
pars ganglionaris into the pars jugularis. The anterior foramen is the trigeminal 
foramen, the more posterior the facial foramen. The hyomandibular nerve passed 
through the facial foramen into the pars jugularis and out through the posterior 
opening of the pars jugularis. The anterior opening of the pars jugularis is 
separated from the posterior opening by a narrow splint of bone. The posterior 
opening transmitted the hyomandibular nerve, the jugular vein, and the orbital 
artery. The palatine nerve did not enter the pars jugularis but passed through a 
small foramen medial to the facial foramen down into the myodome. The anterior 
opening of the pars jugularis is a vertically elongated slit-like aperture through which 
passed the jugular vein, the orbital artery, and the remainder of the trigeminal and 
facial nerves. 

The exoccipitals meet above and below the foramen magnum and each contributes 
to the occipital condyle. Laterally the exoccipital has a prominent foramen which 
transmitted the vagus nerve, the opening being slightly overhung by a flange of 
bone. The glossopharyngeal foramen is smaller and just anterior to the vagus 
foramen. Both of these foramina lie within a groove extending from the posterior 
part of the exoccipital to the posterior opening of the pars jugularis on the prootic. 
Lateral to the occipital condyle and immediately below the vagus foramen, two or 
three very small foramina transmitted the occipital nerves. Laterally the exocci- 
pital is deeply excavated into a membranous concave depression which housed 
pharyngeal musculature. On the posterior face of the skull the exoccipital meets 
the supraoccipital dorso-medially and the epiotic dorsally and dorso-laterally. 

The pterotic bears a dorsal crest forming the outer wall of the post-temporal fossa. 
On the lateral face of the neurocranium the pterotic joins the sphenotic and prootic 
in the hyomandibular facet. The more posterior part of the hyomandibular facet 
is made up of the pterotic alone. The otic branch of the infraorbital sensory canal 
passed within the pterotic crest opening anteriorly above the sphenotic. The 
pterotic crest also delimits the dilatator fossa which is an excavation of the dorso- 
lateral surfaces of both the sphenotic and pterotic. 

The intercalar is a small bone, superficial in position, covering part of the suture 
between the exoccipital and pterotic. The intercalar is wrapped around the 



J 3 2 



UPPER CRETACEOUS TELEOSTS 



soc 




Fig. 59. Halec eupterygius (Dixon). Neurocranium in posterior view. 



posterior angle of the lateral face of the neurocranium, and is thus visible both in 
posterior and lateral views. Postero-dorsally the intercalar has a facet for the 
ventral strut of the post-temporal. 

The sphenotic is large and well ossified and forms the postero-dorsal corner of the 
orbit. The sphenotic projects ventro-laterally from beneath the frontal contacting 
the pterotic posteriorly and the prootic ventrally. The epiotics appear on both 
dorsal and posterior surfaces of the neurocranium and are separated medially by the 
supraoccipital. The epiotic enters into the composition of the post-temporal fossa 
forming part of the medial wall and floor. The upper limb of the post-temporal 
articulates with the dorsal surface of the epiotic, medial to the posterior end of 
the post-temporal fossa. 

Hyopalatine bones. The hyopalatine bones are shown in medial view in Text- 
figure 60. The hyomandibular is large with a double head. The opercular process 
is ill-defined and approximately midway down the rear edge of the bone. The 
lateral face of the hyomandibular has a prominent lateral crest arising from the 
anterior head region and passing postero- ventrally to the rear edge of the bone. 
The preoperculum rests against the posterior face of this crest. The crest gradually 
shallows and disappears near to the ventral extremity of the bone. The anterior 
extent of the hyomandibular is thin and membranous and covered laterally by the 
metapterygoid. The hyomandibular tapers ventrally where it is circular in cross- 
section. The hyomandibular branch of the facial nerve entered the bone through 
an antero-dorsally directed foramen just below the head. Within the bone the 
nerve divided ; the opercular nerve came out near to the opercular process on the 
rear edge of the bone ; the hyoidean and mandibular nerves passed through foramina 
below the opercular process and continued ventrally on the posterior edge of the 
hyomandibular. 



ESPECIALLY MYCTOPHOIDS 



133 



The symplectic is small and inclined forwards ventrally. Dorsally it is separated 
from the hyomandibular and ventrally lies in a groove on the posterior part of the 
medial face of the quadrate. 

The quadrate is a large triangular bone with a transversely orientated condyle at 
its ventral angle. The groove which housed the symplectic is long and shallow and 
ends just behind the condyle. The postero-dorsal corner of the quadrate is emargi- 
nated. The posterior edge however is thickened and slightly grooved to house the 
leading edge of the preoperculum. Dorsally the quadrate meets the metapterygoid 
and anteriorly the ectopterygoid. 



.hm.h 



sop 




Fig. 60. Halec eupterygius (Dixon). Hyopalatine and opercular bones and mandible 

of the left side, medial view. 

The metapterygoid is large, thin and membranous. On its lateral face it has a 
crest which originates antero-ventrally and passes postero-dorsally. More dorsally 
this crest marks the anterior edge of the metapterygoid. 

The ectopterygoid is bent through a slight angle posteriorly where it extends 
below the anterior part of the quadrate. The anterior end of the ectopterygoid 
abuts against the palatine and the dorsal edge contacts the endopterygoid. Teeth 
are present on the ectopterygoid and are of two types. The first are prominent 
marginal teeth which decrease in size posteriorly where they also become closely 
crowded. These teeth are acutely pointed, recurved and with a slight lateral 



'34 



UPPER CRETACEOUS TELEOSTS 



compression. Fine longitudinal striations are present on the teeth and on the 
expanded bases. The second type of teeth on the ectopterygoid are minute and 
collected into a patch posteriorly. The ectopterygoid meets the palatine anteriorly 
in an elongated suture, extending forwards alongside the palatine. 

The endopterygoid is thin and membranous, meeting the metapterygoid poster- 
iorly. The ventral surface of the endopterygoid is convex and supports a large oval 
tooth-patch. 

The palatine is prominent and bears three rows of large teeth, identical in form to 
those on the ectopterygoid. Posteriorly the palatine tapers to fit between the 
ectopterygoid and endopterygoid. At the posterior end the teeth on the palatine 
are reduced to a single row. The tooth row on the ectopterygoid is continuous with 
the most lateral tooth row on the palatine. In this outer tooth row the teeth 
decrease in size anteriorly and posteriorly. The central row of teeth are smaller 
and more irregular both in size and position. The innermost tooth row extends the 
entire length of the palatine and has the largest teeth in the upper jaw. The dorsal 
surface of the palatine is longitudinally grooved for the reception of the ventral 
expansions of the lateral ethmoid. Anteriorly the palatine is associated with the 
lateral edge of the vomer. 



Dermal upper jaw. The dermal upper jaw is shown in lateral view in Text-figure 
61. The premaxilla is elongate and forms the majority of the oral border. The 
anterior region is expanded where it wraps around the front end of the snout to 
contact its partner of the other side. This expanded region is ornamented with 
numerous, evenly spaced, minute tubercles. Posteriorly the premaxilla is long and 




art ang 

Fig. 6i. Halec eupterygius (Dixon). Restoration of the skull in lateral view. 



ESPECIALLY MYCTOPHOIDS 135 

slender, lying beneath the maxilla. The complete oral edge of the premaxilla has a 
fringe of minute conical teeth. 

The maxilla is also long and thin and along most of its length is closely associated 
with the premaxilla. The anterior part of the maxilla curves inwards and this 
medial region is associated with a very small maxillary process on the dorsal surface 
of the palatine. Anteriorly the maxilla also articulates with the vomer and the 
mesethmoid. The maxilla enters the gape posterior to the premaxilla, and supports 
a row of teeth. These teeth are enlarged, well spaced and inclined forwards. 

The maxilla supports a single supramaxilla which is elongate and pointed 
anteriorly. The lateral face of the supramaxilla is ornamented with radiating 
tuberculated ridges. 

Mandible. The mandible is shown in medial and lateral views in Text-figures 60 
and 61. The dentary forms the greater part of both dorsal and ventral edges of 
the mandible, and is constricted at the symphysis. The oral margin of the dentary 
is produced into a medial tooth bearing flange. The teeth are arranged in two rows, 
a marginal row of smaller teeth and an inner row of larger ones. These teeth are 
identical in form to those on the palato-pterygoid. 

The articular forms the posterior part of the dorsal border of the mandible as well 
as part of the ventral border and lateral face. The articular facet is concave and 
transversely orientated. The articular extends anteriorly on to the medial face of 
the dentary. The mandibular sensory canal ran within a tube in the ventro-lateral 
region of both the articular and dentary. The lateral face of the mandible is smooth 
dorsally, but ventrally is ornamented with small tubercles. The angular is a small 
knob of bone attached to the postero-ventral region of the articular. 

Opercular bones. The opercular bones are shown in medial and lateral views in 
Text-figures 60 and 61. The preoperculum is long and narrow with a slight forward 
curvature ventrally. The upper limit of the bone lies behind the hyomandibular 
head whilst the anterior edge rests against the rear edges of both the hyomandibular 
and the quadrate. The preopercular sensory canal ran in a groove on the lateral 
face of the bone. The flange overhanging the groove is ornamented with ridges and 
tubercles. This ornamentation continues on the lower end of the bone. 

The operculum has an irregularly concave ventral edge. From the articular 
facet an internal horizontal strengthening ridge crosses the operculum to terminate on 
the posterior edge. The lateral face of the operculum is ornamented with minute 
tubercles radiating outwards from the region of the opercular facet. The sub- 
operculum is large and deep with its dorsal edge covered by the ventral part of the 
operculum. Ventrally the suboperculum is thin and membranous and has an 
ornamentation of bony tubercles. The interoperculum lies medial to the ventral 
part of the preoperculum, and has a similar ornamentation to the other opercular 
bones. 

Pectoral girdle and fin. The pectoral girdle is shown in medial view in Text- 
figure 62. The post-temporal has a large, plate-like upper limb articulating with the 
epiotic, and a shorter strut-like ventral limb articulating with the intercalar. The 
supracleithrum articulated with the posterior region of the post-temporal, and the 



136 



UPPER CRETACEOUS TELEOSTS 



lateral line sensory canal passed through the head of the supracleithrum and the 
post-temporal. The flattened dorsal surface of the post-temporal is ornamented 
with raised bony tubercles. 

The supracleithrum is expanded and overlaps the upper end of the cleithrum. 
The lateral face of the supracleithrum is ornamented with a few weak ridges. The 
cleithrum is sigmoid in shape with its anterior edge produced medially. The endo- 
skeletal girdle attaches to the internal face of the cleithrum. 

The scapula contacts the cleithrum laterally, the coracoid ventrally and the 
mesocoracoid antero-medially. A large scapular foramen perforates the centre of 
the bone. Postero-dorsally the scapula has a prominent facet with which the 
anterior fin ray articulated. The remainder of the posterior edge is grooved. The 
coracoid is much larger than the scapula and considerably expanded antero-ventrally. 

d.pt 
P. 1 



SCO 



sea. for 




cor 



10mm 



Fig. 62. Halec euplerygius (Dixon). Pectoral girdle of the left side in medial view. 



ESPECIALLY MYCTOPHOIDS 137 

The cleithrum and coracoid meet at their anterior extremities, leaving a fenestra 
between the two bones. Posteriorly the coracoid projects backwards below the fin 
insertion. The mesocoracoid arch is relatively narrow. 

Vertebral column. The vertebral column consists of 45 vertebrae of which at 
least 20 and possibly as many as 22 are caudal. Each centrum is slightly longer 
than deep and is mesially constricted. The anteriormost neural arches are laterally 
compressed and expanded, not fused to the corresponding centra. More posteriorly 
the neural arches are more intimately connected to the centra. The precaudal 
vertebrae bear ventro-lateral transverse processes which support the pleural ribs. 
The ribs are long and narrow and decrease in size towards the end of the precaudal 
region. The first three or four caudal vertebrae have laterally flattened haemal 
arches without prominent spines. Haemal arches posterior to these are less 
expanded and are drawn out into backwardly curved haemal spines. 

Halec haueri (Bassani) 

(Text-figs. 63, 64) 

1879 Elopopsis haueri Bassani : 164. 
1946 Halec haueri (Bassani) ; d'Erasmo : 90. 
d'Erasmo lists the earlier references. 

Diagnosis (emended). Halec species of standard length not exceeding 22 cm. 
Length of head with opercular apparatus greatly exceeds the maximum depth of 
the trunk, which is contained almost four times in the length from the pectoral arch 
to the base of the caudal fin. Length of the mandible exceeding the depth of the 
head at the occiput. Vertebrae 40 in number of which 18 are caudal. Dorsal fin 
with 13 or 14 rays, anal fin small and remote with 9 feeble rays. 

Holotype. Almost complete but badly weathered specimen in the Geologische 
Bundesanstalt, Vienna, from the Lower Cenomanian of Lesina (= Hvar) in the 
Adriatic. 

Material. The holotype and several further specimens in the Naturhistorisches 
Museum, Vienna. The major part of the description was obtained from a single 
excellent specimen, number 1902. 11. 5, in the Bayerische Staatssammlung fur 
Palaontologie, Munich. All of the material is from the Lower Cenomanian of 
Lesina. 

Description. Skull. The entire head region differs little from that of Halec 
eupterygius already described. The skull-roof is remarkably similar with the same 
arrangement of tubercular ornamentation, the enormous extent of the frontals, the 
unroofed post-temporal fossa and the prominent supraorbital bone. The jaw 
suspensorium is vertical with the quadrate condyle lying below the occiput as in 
Halec eupterygius. The dermal upper jaw is toothed, the maxilla bearing the 
characteristic forwardly inclined teeth at the posterior end. 

Certain of the bones of the hyoid arch are shown whereas they were not preserved 
in Halec eupterygius. Both the ceratohyal and epihyal are long and shallow and 



«3« 



UPPER CRETACEOUS TELEOSTS 







a 
o 

+-> 

% 
en 

D 

A 



a 

o 

s 

en 



C 

en 
en 

PQ 






CO 
vO 






ESPECIALLY MYCTOPHOIDS 



139 



support at least 15 branchiostegal rays. The first 8 rays are small and of uniform 
shape, the remaining 7 gradually increase in size until the terminal ones are large 
and flattened. The urohyal is long and formed of two laminae of bone connected 
dorsally. 

Paired fins and girdles. The pectoral girdle is like that of Halec eupterygius, but 
in addition a large postcleithrum projects postero-ventrally from the lower end of the 
supracleithrum. Eleven or 12 rays make up the pectoral fin, the first is unbranched, 
stout and basally ornamented with small tubercles. The pelvic bones lie below the 
dorsal fin and the fins are as large as the pectorals, also consisting of approximately 
n rays. Most of the rays are branched and ornamented in their basal regions. 

Vertebral column. The vertebral column is shown in the restoration of the entire 
skeleton in Text-figure 63. It consists of approximately 40 vertebrae, of which 18 
are caudal. The centra are slightly longer than deep and mesially constricted. 
In the precaudal region the neural arches are easily displaced from the centra 
indicating the looseness of the attachment in life. Ventro-lateral transverse pro- 
cesses occur on the precaudal centra for the articulation of the long, thin, flattened 
ribs. The first two caudal vertebrae only possess haemal arches with which the 
ribs articulate. Both epineurals and epipleurals are present on the first 30 vertebrae. 

Median fins and tail. The median fins are shown in the restoration, Text-figure 
63. The dorsal fin consists of 13 or 14 rays and is just within the anterior half of 
the back. The first finray is short, stout and unbranched with ornamentation on 




Fig. 64. Halec haueri (Bassani). Caudal fin skeleton in lateral view. 



i 4 o UPPER CRETACEOUS TELEOSTS 

its basal region. The remaining rays are branched but unornamented. The 
anteriormost radial is expanded into a large anterior keel. The remainder of the 
radials are unexpanded. 

The anal fin is small and remote with 9 feeble rays. The first ray is unbranched 
and unornamented and none of the radials is expanded. 

The caudal skeleton (Text-fig. 64) is made up of six vertebrae, four preural 
vertebrae and two ural vertebrae. Preural vertebrae 3 and 4 bear normal neural 
and haemal spines, although the latter are somewhat expanded. The neural and 
haemal spines of these preural vertebrae support the accessory rays of the caudal 
fin. Preural vertebra 2 does not possess a neural spine, simply an expanded neural 
arch. Preural vertebra 1 and ural vertebra 1 are fused together to produce a 
compound centrum. The haemal spine of preural vertebra 1 (the parhypural), 
and the first two hypural elements articulate with the ventral edge of this compound 
centrum. Ural vertebra 2 is present as a small centrum supporting the hypurals 
of the upper lobe of the caudal fin, of which there are at least three. A large rod- 
like uroneural, inclined postero-dorsally, is associated with the dorso-lateral parts 
of both the ural vertebrae. The uroneural appears to be expanded anteriorly in 
the mid-line above the first preural vertebra, and is considered to be a stegural. 
A single epural is visible above the anterior expansion of the uroneural. 

The caudal fin consists of 17 branched principal fin rays, together with at least 
16 accessory fin rays. 

Scales. The entire body is covered with a uniform coat of thin, minute scales. 
The scales along the course of the lateral line are enlarged and bear a posteriorly 
directed spine from the lateral face. The lateral line scales become progressively 
larger towards the caudal fin and on the caudal peduncle are irregular in shape with 
an enormous spine. 



Genus PHYLACTOCEPHALUS Davis, 1887 

Diagnosis (emended). Halecidae in which the head is short but deepened ; 
body also shortened and deepened. Jaw suspensorium vertical. Supraorbital 
absent. Approximately 10 branchiostegal rays. Vertebrae 32 in number. 
Dorsal fin in second quarter of the back. Pelvics sub-thoracic, larger than pectorals. 
Epineurals on the first 19 vertebrae. Complete covering of minute scales, enlarged 
along the lateral line. 

Type species. Phyladoccphalus microlepis Davis. 

Remarks. This genus was erected by Davis (1887) but subsequently Woodward 
(1901 : 218) synonymised it with Halec. There are sufficient differences to warrant 
separation into two distinct genera and the original generic name of Phylactocephalus 
is re-erected. 



ESPECIALLY MYCTOPHOIDS 141 

Phylactocephalus microlepis Davis 
(Text-figs. 65, 66) 

1887 Phylactocephalus microlepis Davis : 605, pi. 35, fig. 2. 
1901 Halec microlepis (Davis) Woodward : 218. 

Diagnosis (emended). Phylactocephalus of standard length not exceeding 18 cm. 
Length of the head with the opercular apparatus just exceeds the maximum depth 
of the trunk. Length of the mandible equivalent to the depth of the head at the 
occiput. Vertebrae 32 in number of which 16 are caudal. Dorsal fin with 17 rays 
occupying the second quarter of the back. Anal fin small and remote with 11 rays. 

Holotype. Specimen number P. 4757 in the B.M.N.H., a complete but flattened 
fish from the Middle Cenomanian of Hajula, Lebanon. 

Material. Holotype and several other specimens in the B.M.N.H., numbers 
P. 105, P. 46453, P. 47318 from the Middle Cenomanian of Hakel ; P. 47329 from the 
Middle Cenomanian of Hajula, Lebanon. 

Remarks. The type specimen was initially poorly described by Davis (1887 : 
605, pi. 35, fig. 2). His description was later corrected by Woodward (1901 : 218) 
when he transferred the species to the genus Halec. The type specimen was 
developed in acetic acid by the transfer method together with a complete, exception- 
ally well preserved specimen, number P. 105. Since all the material is flattened, 
neurocranial description is difficult, however the postero-lateral region of the skull- 
roof is partially shown in specimen number P. 47318. 

Description. Neurocranium. The majority of the skull-roof is formed from 
the enlarged frontals which extend posteriorly almost to the occipital border. 
Laterally the frontal is curved above the orbital region to overlap the dorsal surface 
of the sphenotic and form the hind end of the orbit. Postero-laterally the frontal 
is deeply indented, and this marks the anterior extent of the unroofed post-temporal 
fossa. The frontals are uniformly ornamented with minute bony tubercles. In the 
snout region the frontals curve antero-ventrally. The sphenotic projects from 
beneath the frontal at the rear end of the orbit. The pterotic forms the postero- 
lateral border to the neurocranial roof, contacting the frontal and parietal medially. 
The dorsal surface of the pterotic has a prominent longitudinal crest which separates 
the post-temporal fossa from the dilatator fossa. The crest has a small posterior 
foramen which transmitted the lateral line canal into the pterotic. Ventro-laterally 
the pterotic has an elongated groove and this represents the hind region of the 
hyomandibular facet. Anteriorly the hyomandibular facet is cup-shaped and 
formed of sphenotic and prootic. 

Within the orbit the parasphenoid is slightly curved and possesses small ventro- 
laterally inclined flanges. Towards the snout the parasphenoid flattens where it 
contacts the vomer. The vomer appears to contact the underside of the mesethmoid 
anteriorly and has slight lateral extensions which parallel those of the mesethmoid. 

Hyopalatine bones. The hyomandibular is large and has a double head. The 
posterior part of the head is elongate and lies in a horizontal plane but the anterior 



142 UPPER CRETACEOUS TELEOSTS 

part is short and inclined antero-ventrally. The opercular process is pronounced 
and in the upper half of the bone. The lateral face of the hyomandibular bears a 
prominent crest running from the anterior head region back to the posterior edge. 
In front of the crest the hyomandibular is thin and tapers ventrally, this region is 
covered laterally by the metapterygoid. 

The quadrate is a stout triangular bone with a transversely arranged condyle. 
Postero-dorsally the upper edge is excavated. This indentation marks the upper 
region of the groove on the medial face of the quadrate which housed the symplectic. 
The symplectic is narrow and inclined forwards ventrally. The posterior edge of 
the quadrate is thickened. Dorsally the quadrate contacts the metapterygoid and 
anteriorly attaches to the postero-medial region of the ectopterygoid. 

The metapterygoid is a flattened membranous bone overlapping the anterior 
extent of the hyomandibular. Laterally on the metapterygoid a shallow crest 
passes obliquely backwards. More dorsally this crest marks the anterior edge of the 
metapterygoid. Anteriorly the metapterygoid contacts the postero-dorsal part of 
the ectopterygoid. 

The ectopterygoid is elongate and bent through a slight angle near to its posterior 
end. Anteriorly the ectopterygoid meets the palatine in an elongate suture. A 
single row of teeth is present on the ectopterygoid, and these are acutely pointed, 
laterally compressed and evenly spaced. The teeth decrease in length towards the 
posterior region, where they are inclined slightly forwards. The teeth are finely 
striated, unexpanded basally and firmly fused to the bone. The tooth row is 
continuous on to the palatine. As well as the marginal row of teeth the palatine 
has an inner row of longer, more needle-like teeth. The palatine is supported 
by the lateral ethmoid dorsally. This latter bone is expanded above the dorso- 
medial surface of the palatine and firmly attached to it. Anteriorly the palatine 
articulates with the lateral edge of the vomer. The palatine has a small dorso- 
lateral maxillary process near its anterior end. This process is excavated terminally, 
and was filled and capped by cartilage in life. 

Dermal upper jaw. The premaxilla is elongate and makes up the anterior half of 
the oral border of the upper jaw. Anteriorly the premaxilla is deepened and 
wrapped around the snout. The lateral face is ornamented with a large number of 
minute tubercles. The premaxilla tapers posteriorly and lies ventral to the maxilla. 
The dorsal edge of the premaxilla is inserted into a shallow groove on the ventro- 
lateral region of the maxilla. A fringe of small pointed teeth is present on the oral 
border of the premaxilla. 

The maxilla is long and thin with an expanded and medially inclined head. The 
facet on the maxilla which housed the maxillary process of the palatine is small and 
bounded fore and aft by small bony projections. A marginal tooth row is con- 
tinuous with that on the premaxilla, but posteriorly the teeth are lengthened and 
inclined forwards. 

Above the posterior end of the maxilla is a conspicuous supramaxilla which 
narrows anteriorly but is deeper posteriorly where it extends beyond the end of the 
maxilla. The supramaxilla is ornamented with minute tubercles. 



ESPECIALLY MYCTOPHOIDS 143 

Mandible. The dentary forms the major part of the mandible and has a V- 
shaped posterior indentation. The symphysial region is considerably constricted 
but posteriorly the dentary is deep. The oral border supports a dense cluster of 
small acutely pointed teeth. Whether these teeth were arranged in definite rows 
could not be ascertained, although there is a distinct row of minute conical teeth 
on the extreme oral edge. This marginal row of teeth resembles that on the 
dermal upper jaw, whereas the more internal teeth are similar to those on the palato- 
pterygoid. 

The articular forms the posterior part of the mandible and has a shallow, trans- 
versely orientated facet. On the postero-lateral surface a small pore, just below 
and behind the articular facet, transmitted the mandibular sensory canal. The 
sensory canal ran within the articular and opened into a deep groove anteriorly. 
This groove is continued on the ventro-lateral part of the dentary. Along much of 
its length on the dentary the groove is covered by a dorsal flange of bone. 

The lateral surface of the mandible is ornamented with minute tubercles. Above 
the sensory canal groove the tubercles radiate backwards from the symphysis, but 
below the groove the tubercles are much less prominent and regular. The articular 
ornamentation is confined to the region below the articular facet. 

The angular is a small knob of bone attached to the postero-ventral region of the 
articular and is strongly ornamented with bony tubercles. 

Opercular bones. The preoperculum is deep and narrow and vertical in position. 
Dorsally the preoperculum ends below the upper limit of the hyomandibular. The 
anterior edge of the preoperculum lies against the posterior face of the crest on the 
hyomandibular and ventrally abuts against the posterior edge of the quadrate. 
The ventral extremity of the preoperculum is slightly broadened. The preopercular 
sensory canal ran within a groove in the upper half of the bone, but in a definite tube 
more ventrally. This tube opens anteriorly at the lowermost extremity of the 
preoperculum beneath the quadrate condyle. The lateral face of the preoperculum 
is liberally ornamented with minute tubercles. 

The operculum has a straight, vertical anterior edge and the opercular facet 
occurs approximately one-third of the way down. Dorsally and posteriorly the 
operculum is rounded, but ventrally the edge is obliquely inclined. From the 
articulatory facet a horizontal strengthening ridge extends back across the inner face 
of the operculum. The operculum is ornamented with bony tubercles which 
radiate from the region of the opercular facet. The tubercles are absent from a 
narrow region anteriorly and antero-dorsally. 

The suboperculum is large and deep with its dorsal edge lying medial to the 
operculum. From the antero-dorsal region of the suboperculum a small projection 
extends dorsally. The whole anterior edge of the suboperculum is thickened but 
the remainder, like the operculum, is thin. Externally the characteristic tubercles 
are arranged randomly. 

A small, ornamented interoperculum is present beneath the ventral extremity of 
the preoperculum and suboperculum. 



i 4 4 



UPPER CRETACEOUS TELEOSTS 



Hyoid arch and branchiostegal rays. Separate dorsal and ventral hypohyals are 
attached to the anterior end of the shallow, elongate ceratohyal. The epihyal, like 
the ceratohyal, is long and shallow. Both ceratohyal and epihyal bear branchio- 
stegal rays ventrally. The number of branchiostegal rays is approximately 10 and 
these become progressively stouter posteriorly. The urohyal is elongate and formed 
of two narrow lamellae of bone which contact each other along their dorsal edges. 
The urohyal extends back to the level of the articular facet of the mandible. 



v.pci 




, 25mm , 
Fig. 65. Phylactocephalus microlepis Davis. Pectoral girdle of the right side in lateral 



Pectoral girdle and fin. The pectoral girdle is shown in lateral view in Text-figure 
65. The post-temporal has a flattened, oval, dorsal limb articulating with the 
epiotic, and a shorter, strut-like ventral limb articulating lower down on the posterior 
face of the neurocranium. Posteriorly the post-temporal is expanded to provide a 
medial facet for the articulation of the supracleithrum. The dorsal surface of the 
upper limb of the post-temporal is ornamented with minute tubercles and the 
lateral line sensory canal passed through the main body of the bone. 

The supracleithrum is laterally flattened, with a thickened anterior edge. It is 
ornamented with minute tubercles arranged in rows following the long axis of the 
bone. The lateral line canal entered the supracleithrum midway down its posterior 



ESPECIALLY MYCTOPHOIDS 145 

edge. Ventrally the supracleithrum attaches to the lateral face of the upper region 
of the cleithrum. 

The dorsal postcleithrum is similar in form to the supracleithrum, with a thickened 
anterior edge. The ornamentation of tubercles on the lateral face of the dorsal 
postcleithrum is from two distinct areas and the posterior edge of the bone is 
indented near its mid-point. (This seems to indicate that the dorsal postcleithrum 
represents the fusion of two separate elements.) The ventral postcleithrum is a 
small flattened rod of bone attached to the antero-ventral part of the dorsal post- 
cleithrum. 

The cleithrum is large and curved with its anterior edge inclined medially. 
Dorsally the cleithrum is narrow but more ventrally it expands where it follows the 
contour of the edge of the suboperculum. The anterior termination of the cleithrum 
is medial to the lowermost extremity of the preoperculum. The cleithrum is 
ornamented with radiating rows of tubercles. 

The endoskeletal girdle attaches to the internal face of the cleithrum and projects 
posteriorly and medially. The posterior edge of the scapula is thickened and has 
a distinct saddle-shaped facet for the anterior fin ray. The coracoid is larger and 
curves forwards ventrally to contact the anterior extremity of the cleithrum. A 
strengthening ridge from the scapulocoracoid junction passes forwards along the 
dorsal surface of the anterior extension of the coracoid. The coracoid projects 
posteriorly below the fin insertion. 

The pectoral fin has approximately 12 rays, the first ray being unbranched. All 
of the rays are stout basally and some are ornamented with bony tubercles. The 
fin is inserted low on the flank. 

Pelvic girdle and fin. Each pelvic bone is stout, flattened and with a prominent 
vertical strengthening ridge running along its length. The bones taper anteriorly 
to contact each other at the extremities. The posterior regions are thickened to 
form a prominent articulatory region for the fin rays and unite in the mid-line. 

The pelvic fin consists of approximately 11 rays and the first ray is unbranched. 
As in the pectoral fin the rays are stout, and certain of them are ornamented with 
ridges and tubercles. The rays are exceedingly elongate, the fin itself being much 
larger than the pectoral. The pelvic fins are in a sub-thoracic position, but the 
girdle does not contact either the cleithrum or the postcleithrum. 

Vertebral column. The vertebral column is shown in the restoration, Text-figure 
66. It has approximately 32 vertebrae, of which 16 are caudal. The individual 
centra are slightly longer than deep and are mesially constricted. The anterior 
neural arches are not intimately connected with the centra but the more posterior 
arches have a much closer connection, although the suture between the two structures 
is always clearly visible. In the precaudal region paired ventro-lateral transverse 
processes are found on each centrum and associated with these are long, thin, 
flattened pleural ribs. The first two caudal vertebrae bear haemal arches but only 
rudimentary haemal spines. More posteriorly the haemal spines are fully formed. 
Epineurals and epipleurals are present on the first 23 vertebrae. 



146 UPPER CRETACEOUS TELEOSTS 

Median fins and tail. The median fins are shown in the restoration, Text-figure 
66. The dorsal fin has 17 stout rays, and occupies the second quarter of the back. 
The first ray is shorter than the subsequent ones and unbranched, with its basal 
region ornamented. The remainder of the rays are branched and supported by 
elongate proximal radials. The anteriormost proximal radial is expanded anteriorly 
into a large keel-like structure. Traces of the medial radials are visible extending 
between the bases of adjacent fin rays. 

The small anal fin consists of 11 feeble rays and is remote in position. The first ray 
is short and unbranched, the remainder are all branched. None of the proximal 
radials is expanded. 

The caudal fin is supported by four or five centra. Three of the centra are free 
preural centra, the first preural vertebra appearing to be fused with the centrum of 
ural vertebra 1. Ural vertebra 2 may be present as a terminal half-centrum. The 
remainder of the caudal skeleton is indeterminable. 

The fin itself is composed of 17 branched principal rays, and approximately 16 
accessory rays. 

Scales. The entire body is covered with a uniform coat of small scales, regular in 
shape and arrangement and closely packed. The scales are not present on the 
opercular bones but do cover most of the cheek region. Each scale bears a minute, 
longitudinal crest and these crests are continuous, producing a regular longitudinal 
series of parallel ridges along the body. The scales along the course of the lateral 
line are enlarged and thickened with the longitudinal crests becoming much more 
pronounced towards the caudal peduncle. On the peduncle the crests fuse with 
one another to produce a lateral, longitudinal flange. 



Genus HEMISAURIDA Kner, 1867 

Diagnosis (emended). Halecidae in which the head is not deepened but slightly 
elongated ; body shallow and not elongated. Supraorbital absent. Jaw sus- 
pensorium inclined anteriorly. Preoperculum with a greatly extended posterior 
spine. 16 branchiostegal rays. Dorsal fin in the second quarter of the back. 
Pelvics abdominal but smaller than the considerably elongated pectorals. Epi- 
neurals on the first 23 vertebrae. Body completely naked. 

Type species. Hemisaurida neocomiensis Kner. 

Remarks. The genus was erected by Kner in 1867 to contain the single species 
Hemisaurida neocomiensis. The holotype of this species is a fragment of the head in 
the Geologische Bundesanstalt, Vienna. The specimen is from the Lower Cenomanian, 
Comen, near Trieste. A more complete specimen of this species, lacking the tail 
region and the tip of the snout is also present in Vienna. These are the only speci- 
mens of the type species and both were examined. 



ESPECIALLY MYCTOPHOIDS 



147 




M 






<[> 




& 






T) 




V 




+J 




+-> 


> 


e 


Q 






8 

5N 



£ 

&H 






i 4 8 UPPER CRETACEOUS TELEOSTS 

Hemisaurida hakelensis sp. nov. 

(Text-figs. 67-69) 

Diagnosis. Hemisaurida of standard length up to 5-5 cm. Length of the head 
with the opercular apparatus two and a half times the maximum depth of the head, 
and equal to just under half the length from the pectoral arch to the base of the 
caudal fin. Vertebrae 35 in number of which 17 are caudal. Dorsal fin in the 
second quarter of the back with n rays. Anal fin remote and consisting of 9 rays. 

Holotype. B.M.N.H. number P.48777, a flattened fish from the Middle Ceno- 
manian, Hakel, Lebanon. 

Material. The holotype and specimen numbers P. 48778, P. 48779, P. 48780 in 
the B.M.N.H., from the Middle Cenomanian of Hakel, Lebanon. 

Description. Neurocranium. The neurocranium is somewhat elongate, shallow 
and with a large orbital region. The orbit is bordered dorsally by the frontal and 
limited anteriorly by a prominent lateral ethmoid. At the hind end of the orbit, the 
frontal extends laterally above the sphenotic. The sphenotic passes ventro- 
medially to contact the prootic within the anterior portion of the hyomandibular 
facet. The prootic forms part of the lateral wall of the cranial cavity and both the 
anterior and posterior openings of the trigemino-facialis chamber can be seen in it. 
The openings are separated by a strut of the prootic. Ventrally the prootic contacts 
the ascending process of the parasphenoid. At the base of the ascending process is 
a notch in the anterior edge marking the passage of the efferent pseudobranchial 
artery. Posterior to this notch, still in the parasphenoid, is a forwardly directed 
foramen which transmitted the internal carotid artery. The parasphenoid, in the 
base of the orbit, bears lateral flanges. 

Hyopalatine bones. The hyomandibular is a shallow bone with an elongated head. 
The anterior region of the head fits into the cup-shaped facet formed from the 
sphenotic and prootic, whilst the posterior region fits into an elongated facet 
within the pterotic. A prominent lateral crest, from the anterior head region, 
passes postero-ventrally to the rear edge of the hyomandibular in the vicinity of 
the opercular process. The crest passes ventrally along the rear edge of the bone 
and disappears at the ventral extremity. A second crest passes ventrally from 
the posterior head region of the hyomandibular to contact the main crest where 
the latter meets the posterior edge of the bone. The hyomandibular curves strongly 
forwards ventrally (Text-fig. 67). The anterior part of the hyomandibular is thin 
and laminate and overlapped by the prominent metapterygoid. This latter bone 
supports an oblique crest and is relatively large, with rounded posterior and dorsal 
edges. The ventral edge is straighter and contacts the dorsal edge of the quadrate. 
The quadrate is large and triangular with a prominent condyle. The rear edge of the 
quadrate forms a continuation of the forward curvature of the ventral hyomandi- 
bular region. The quadrate condyle thus lies beneath the rear end of the orbit. 
The ectopterygoid and the palatine form a continuous strut of bone ascending 
gradually from the quadrate. The suture between the two bones is below the 
extreme anterior orbital region. The endopterygoid is thin and attached to the 



ERRATUM 



p. 148. Under Holotype delete all and replace by: 

American University, Beirut, number 104710, 104786, a complete fish in counter- 
part from the Middle Cenomanian, Hakel, Lebanon. 

under Material delete all and replace by: 

The holotype and another specimen in counterpart, American University, Beirut, 
number 106805, 106819, from the Middle Cenomanian of Hajula, Lebanon. 



ESPECIALLY MYCTOPHOIDS 149 

medial edge of the ectopterygoid, lying in the roof of the mouth associated with the 
ventro-lateral flange of the parasphenoid. Anteriorly the palatine articulation with 
the mesethmod region of the skull is unexpanded. The lateral ethmoid contacts 
the postero-dorsal surface of the palatine. The palato-pterygoid bar bears a 
continuous, single series of teeth along its entire length. 




mpt' q "- N ps ,op si)p cl —r 

Fig. 67. Hemisaurida hakelensis sp. nov. Restoration of the skull in lateral view. 

Dermal upper Jaw. The dermal upper jaw is shown in lateral view in Text-figure 
67. The premaxilla is elongate and very shallow with no ascending process. 
Anteriorly the premaxilla is slightly expanded and wrapped around the snout. The 
premaxilla forms approximately the anterior half of the oral border of the upper jaw, 
and supports a fringe of minute teeth. The maxilla is almost twice as long as the 
premaxilla and is narrow along its entire length. It enters the oral border of 
the jaw behind the premaxilla and bears teeth on its ventral edge. The maxillary 
teeth are more prominent than those on the premaxilla and increase in size poster- 
iorly, where they are forwardly inclined. At its extreme anterior end the head of 
the maxilla curves in towards the mid-line and bears a medial facet which is asso- 
ciated with the dorsal palatine surface. Above the maxilla is a long, low supra- 
maxilla which just extends beyond the posterior end of the maxilla. 

Mandible. The mandible is shown in lateral view in Text-figure 67. It is long, 
shallow and formed mostly of dentary. The articular facet is shallow and not well 
demarcated and there is no retroarticular process. The lateral face of the articular 
has a flange passing forwards horizontally from the articular facet. The groove and 
shelf produced above this flange support the maxilla and supramaxilla when the 
jaws are closed. The oral margin of the dentary has a series of small, clustered 
teeth. 

Opercular bones. The opercular bones are shown in lateral view in Text-figure 
67. The preoperculum is deep and narrow and curves forwards ventrally. Postero- 
ventrally it is produced into a prominent acutely pointed spine. The tube which 
contained the preopercular sensory canal traverses the entire length of the pre- 
operculum and opens ventrally behind the condyle. The operculum is a very large 



150 UPPER CRETACEOUS TELEOSTS 

bone, as long as deep, with a rounded posterior edge. The suboperculum is reduced 
in extent and the interoperculum is small. 

Hyoid arch and branchiostegal rays. Both the ceratohyal and the epihyal are 
visible below the dentary. Each is elongate and shallow, the ceratohyal attaching 
to a hypohyal anteriorly. The elongate median urohyal has an anterior articular 
region which is associated with the ventral region of the hypohyal. Sixteen bran- 
chiostegal rays are present, 8 of which articulate with the ceratohyal and 8 with the 
epihyal. The anteriormost 9 or 10 branchiostegal rays are uniform in length and 
thickness, but the remainder gradually increase in length and girth. 

Paired fins and girdles. Due to the increase in size of the operculum the girdle has 
moved posteriorly and the occipital connection through the medium of the post- 
temporal is lengthened. Although the post-temporal is elongate it is of the 
normal forked nature with a more prominent flattened dorsal limb. The supra- 
cleithrum articulates with the postero-medial part of the post-temporal and from 
here it extends postero-ventrally to pass lateral to the upper part of the cleithrum. 
Two prominent postcleithra form a continuation of the supracleithrum. The 
dorsal post cleithrum is a large oval bone with a thickened anterior edge. The 
ventral postcleithrum is a thin splint of bone attached to the lowermost region of 
the upper. The lateral face of the cleithrum is expanded in a posterior direction 
above the fin insertion. The endoskeletal girdle is attached to the internal face of 
the cleithrum and the coracoid is visible ventrally as the more prominent of the two 
bones. 

The pectoral fin is greatly elongated and composed of approximately 12 rays. 
The first ray is the largest and articulates directly with the posterior edge of the 
scapula. The first ray appears to be unbranched but the remainder gradually 
decrease in length and are all branched. 

The pelvic bones are small slips of bone in the ventral body wall with posterior 
expanded articulatory regions. The pelvic fin consists of approximately 9 rays, the 
first of which is unbranched, the remainder branched. 

Vertebral column. The vertebral column is shown in the restoration, Text-figure 
68. It is composed of 35 vertebrae, of which 17 are caudal. The centra are slightly 
longer than they are deep and mesially constricted. The last few caudal vertebrae 
decrease in length somewhat. The anteriormost neural spines are longer and more 
delicate than those on the caudal region. The anterior neural arches are not rigidly 
attached to the corresponding centra, but the posterior arches are more intimately 
attached. Each neural arch does not extend along the entire dorsal surface of the 
centrum but is confined more to the anterior end, this is much more noticeable in the 
caudal region where the neural arches and spines are larger. Prezygapophyses and 
postzygapophyses are present on every vertebra, but again are more pronounced in 
the caudal region. Small transverse processes are present on the ventro-lateral 
parts of the precaudal centra, and have well developed, long flattened pleural ribs 
articulating with them. The haemal arches gradually extend into haemal spines, a 
complete spine being present by the third caudal vertebra. Pleural ribs still 



ESPECIALLY MYCTOPHOIDS 



151 




0) 



« 



► 
O 

c 



8 






152 



UPPER CRETACEOUS TELEOSTS 



associate with the haemal arches of caudal vertebrae i and 2. 
epipleurals are associated with the first 23 vertebrae. 



Both epineurals and 



Median fins and tail. The median fins are shown in the restoration, Text-figure 
68. The dorsal fin is short and deeper than long, positioned in the second quarter of 
the back. The fin has 12 rays, the first short and unbranched, the remainder 
longer and branched. All of the rays are segmented. The proximal radials are 
long and narrow except for the first which bears an expanded anterior keel. 

The anal fin is remote and is made up of 8 or 9 small feeble rays. The first ray 
appears to be unbranched whilst the remainder are branched. 




2mm 



Fig. 69. Hemisaurida hahelensis sp. nov. Caudal fin skeleton in lateral view. 



The caudal skeleton (Text-fig. 69) consists of five vertebrae, although one of 
these would seem to be a compound vertebra. Three free preural vertebrae are 
present (Preurals 2, 3 and 4), all possessing expanded haemal and neural spines. 
Preural vertebra 1 appears to have fused with ural vertebra 1 and this compound 
vertebra supports the parhypural (haemal spine of preural 1), and two hypural 
elements, the first of which is the larger. The second ural vertebra is present as a 
terminal half-centrum associated with the hypurals to the upper lobe of the caudal 
fin. Two elongate, postero-dorsally directed, rod-like uroneurals are associated with 
the dorso-lateral regions of the compound centrum and the terminal half-centrum. 
One or two small epurals are present anterior to the uroneurals. 



ESPECIALLY MYCTOPHOIDS 153 

Order MYCTOPHIFORMES 
Family SARDINIOIDIDAE nov. 

Diagnosis. Myctophiformes in which the post-temporal fossa is roofed ; the 
parietals meet in the mid-line anterior to the supraoccipital. Maxilla excluded 
from gape but expanded posteriorly with two supramaxillae. Premaxilla with long 
ascending process. Antorbital present, but no supraorbital. Both operculum and 
suboperculum large. Pelvics abdominal with a pelvic splint bone. Fulcral scales 
at base of tail. Vertebral elements incompletely fused. Ural vertebra 2 present 
as a terminal half-centrum. Scales cycloid or feebly ctenoid on body, also present 
on cheeks, operculum, skull-roof and posterior part of the maxilla. 

Genus SARDINIOIDES von der Marck, 1858 

1863 Leptosomus von der Marck : 49. 

1940 Cassandra White and Moy-Thomas : 102. 

Diagnosis (emended). Sardinioididae in which the body is short and somewhat 
deepened. Vertebrae between 30 and 32. Branchiostegals about 9 in number. 
Pleural ribs delicate. Dorsal fin in the front half of the back, with between 10 and 
18 rays. Pectorals small and delicate, situated on the flank. Gap present between 
hypurals 2 and 3 and a corresponding gap between the fin ray bases. Scales ctenoid 
in larger species, cycloid in smaller ones. 

Type species. Sardinioides monasteri (Agassiz). 

Remarks. Hay (1903 : 424) and Arambourg (1954 : 60) have maintained that 
the correct name of this genus is Osmeroides Agassiz, on the grounds that Osmeroides 
monasteri Agassiz is the type species of that genus. However the first publication 
of the name Osmeroides in a binomial was as Osmeroides lewesiensis (Agassiz 1837, 
5 ; pis. 60b, c), an elopoid fish in no way related to Sardinioides, thus preceding the 
first use of Osmeroides monasteri (Agassiz 1839, 5 ; pi. 6od). Sardinioides is there- 
fore the valid generic name for the myctophiform fishes under consideration here. 

Sardinioides minimus (Agassiz) 
(Text-figs. 70-72) 

1839 Clupea minima Agassiz, 5 : 2 : pi. 61, fig. 1. 

1844 Clupea minima Agassiz, 5 ; 2 : 120. 

1866 Clupea minima Agassiz ; Pictet and Humbert : 65. 

1901 Leptosomus minimus (Agassiz) Woodward : 246. 

1903 Osmeroides pontivagus Hay : 424, pi. 33, figs. 1-4. 

Diagnosis (emended). Sardinioides of standard length not exceeding 45 mm. 
Length of head with opercular apparatus just less than one-third of the standard 
length, and equal to the maximum depth of the body. Pectoral fin with 14 rays ; 
pelvic fin with 9 rays. Dorsal fin with 11 rays ; anal fin with 9 rays. Four 
accessory radials in advance of the dorsal fin. Scales mainly cycloid, some feebly 
ctenoid. 



i54 UPPER CRETACEOUS TELEOSTS 

Holotype. At one time in the Amic collection, Paris, but present whereabouts 
unknown. It was a flattened fish from the Middle Cenomanian of Hakel, Lebanon. 

Material. Nine specimens in the B.M.N.H., 6 from the Middle Cenomanian of 
Hakel and 3 from the Middle Cenomanian of Hajula. All 9 specimens were prepared 
by the transfer method in acetic acid. 

Description. Neurocranium. The neurocranium is shallow and reaches its 
maximum depth below the hind end of the orbit. The parasphenoid is long and 
straight. The f rentals form the major part of the skull-roof and extend back to 
cover most of the cranial cavity. They meet in the mid-line in a sinuous suture and 
are widest at the hind end of the orbit where they extend above the dorsal surface of 
the sphenotic. In this region the frontal is indented. Within this indentation the 
dorsal surface of the sphenotic contacts the pterotic. The frontal meets the parietal 
and pterotic posteriorly and the sphenotic and pterotic laterally. The surface of all 
the bones is unornamented. The supraorbital sensory canal ran within the frontal 
above the orbit. Anteriorly the frontal tapers from in front of the orbit to insert on 
the dorsal surface of the broad mesethmoid region. 

The parietals meet in the mid-line of the skull-roof and are rectangular in shape. 
They join the f rentals anteriorly, the pterotics laterally in the roof of the post- 
temporal fossa and the epiotics posteriorly. The supratemporal lies on the posterior 
surface of the parietal. 

The pterotic forms the postero-lateral region of the skull-roof, meeting the frontal 
and parietal medially in the roof of the post-temporal fossa, and the sphenotic 
anteriorly. The lateral surface of the pterotic is deeply excavated to form the 
dilatator fossa and the pterotic contacts the sphenotic in the anterior region of it. 
The hyomandibular facet is composed of sphenotic and pterotic with possibly an 
anterior contribution from the prootic. The otic branch of the infraorbital sensory 
canal passed within the pterotic above the dilatator fossa. The main infraorbital 
sensory canal from the last infraorbital bone connected with the otic branch on the 
dorsal sphenotic region, and then passed into the frontal. 

In all of the specimens in which it is preserved, the last infraorbital remains closely 
attached to the dorsal sphenotic region (as it is in Aulopus, p. 208). Anteriorly the 
neurocranium is indistinct. The mesethmoid is a broad bone bifurcated anteriorly, 
producing two antero-lateral projections which are thickened and incline ventrally. 
The enlarged head of each maxilla fits beneath these projections, being retained in 
position by the vomer below. Ventro-lateral mesethmoid wings are also present 
associated with the palatine. A posterior process of the vomer passes back beneath 
the anterior end of the parasphenoid in the lateral ethmoid region. The lateral 
ethmoids attach to the parasphenoid and palatine ventrally. 

A partially ossified interorbital septum may possibly represent an orbitosphenoid. 
Traces of a basisphenoid pedicel are also evident in the rear of the orbit. 

Lying against the edge of the lateral ethmoid is a fragmentary bone which is 
probably an antorbital. 

Hyopalatine bones. The hyomandibular is upright and narrow with a forwardly 
inclined, single and elongate head. A lateral crest on the anterior region of the head 



ESPECIALLY MYCTOPHOIDS 



155 



passes postero-ventrally to the rear edge of the bone where it continues ventrally. 
At the point of contact with the posterior edge this lateral crest is produced into a 
small spine (Text-fig. 70) . This spine lies lateral to the opercular process (approxi- 
mately one-third of the way down the rear edge of the bone) . The hyomandibular 
tapers ventrally and the symplectic is visible within an indentation of the upper 
edge of the quadrate. The quadrate is stout with a prominent transversely arranged 
condyle. The rear edge is thickened where it contacts the leading edge of the pre- 
operculum. The palato-pterygoid bones were not shown well in any of the specimens 
examined, but they do not appear to differ to any extent from those of Aulopus 
described in the next section (Text-fig. 89, p. 211). 



le as.ppm 

mes / ai-.p.pm 



d.pcl 



vpcl 




mx.p.p 



pm 



Fig. 70. Sardinioides minimus (Agassiz). Restoration of the skull in lateral view. 



Dermal upper jaw. The dermal upper jaw is shown in lateral view in Text-figure 
70. The premaxilla completely excludes the maxilla from the oral border of the 
jaw. Anteriorly there is a prominent, broad ascending process but posteriorly the 
bone becomes long and very thin. The ascending process articulates with the 
anterior mesethmoid region. An articular process from the lateral region of the 
ascending process abuts against a facet on the anterior edge of the maxillary head. 
The alveolar arm of the premaxilla is closely applied to the ventral edge of the 
maxilla for its entire length. 

The maxilla is larger and stouter than the premaxilla and has a medially inclined 
head. The antero-dorsal part has a depression into which the maxillary process 
of the palatine fits. This depression is bounded posteriorly by a flange. The 



156 UPPER CRETACEOUS TELEOSTS 

actual head of the maxilla, in front of this depression, is large and rounded and 
articulates with both the mesethmoid and the vomer. The maxilla becomes 
flattened and deepened and reaches its greatest depth posteriorly. On the dorsal 
edge of the maxilla are two reduced supramaxillae. The anterior is the smaller 
and is pointed in front ; the posterior one has an antero-dorsal projection which lies 
along the dorsal edge of the anterior supramaxilla. All of the dermal upper jaw 
bones are smooth and unornamented. 

Mandible. The mandible is shown in lateral view in Text-figure 70. It is long 
and deepened posteriorly. The dentary forms the greater part of both dorsal and 
ventral edges, as well as the lateral face. On the ventro-lateral part of the dentary 
is a deep groove covered by a dorsal flange. Anteriorly this groove is closed and 
opens to the exterior through one or two large pores. The groove and tube con- 
tained the mandibular sensory canal. The oral border of the dentary bears 
numerous small teeth. 

The articular forms the posterior part of the jaw, with a transversely arranged, 
shallow articular facet. The posterior edge of the bone rises steeply from the 
facet to meet the dentary dorsally as the coronoid process. A small retroarticular 
process occurs behind the articular facet, and bears a lateral groove passing beneath 
the articular facet for the mandibular sensory canal. 

The angular is a triangular knob of bone applied to the posterior end of the 
articular below the retroarticular process. 

Opercular bones. The opercular bones are shown in lateral view in Text-figure 70. 
The operculum is large and deep with a straight, slightly thickened anterior edge, 
a rounded dorsal edge and an obliquely inclined ventral edge. The lateral face is 
unornamented and the opercular facet occurs near the upper limit of the anterior 
edge of the bone. The body scaling is continued on to the operculum. The 
suboperculum is also large and deep with its dorsal edge lying medial to the oper- 
culum. The bone is unornamented but the anterior edge is thickened and extended 
dorsally into a short process, medial to the operculum. The interoperculum is a 
prominent triangular bone lying medial to the ventral preopercular region. It is 
thin and unornamented and ends anteriorly just behind the retroarticular process. 

The preoperculum inclines forwards ventrally where it is considerably broadened. 
It ends dorsally below the head of the hyomandibular, with its leading edge resting 
against the posterior crest of the hyomandibular. Below the hyomandibular crest, 
the anterior edge of the preoperculum is expanded into a slight convexity. In the 
extreme ventral region the leading edge of the preoperculum rests against the 
posterior edge of the quadrate. The preopercular sensory canal ran in a groove with 
a prominent overhanging flange. Body scaling extends onto the posterior regions 
of the preoperculum, but not on to this flange. 

Pectoral girdle and fin. The supratemporal is never well preserved, but a projecting 
posterior portion appears to have housed a sensory canal. The post-temporal 
has a long curved dorsal limb articulating with the epiotic and an equally long 
ventral limb articulating with the posterior face of the neurocranium. The junction 



ESPECIALLY MYCTOPHOIDS 



157 




T3 






X! 



4) 



6fi 



1 
3 






158 UPPER CRETACEOUS TELEOSTS 

of the two limbs is expanded and flattened vertically. The supracleithrum articulates 
with the postero-medial part of the post-temporal. 

The supracleithrum is long and blade-shaped with a thickened anterior edge. 
The cleithrum is large and curved and expanded above the insertion of the fin. The 
endoskeletal girdle is attached to the internal face of the cleithrum, and the coracoid 
extends forwards to contact the cleithrum at the anterior end. Four pectoral fin 
radials support the fin rays and are slightly elongated and mesially constricted. 
Two postcleithra are attached to the rear edge of the cleithrum. The dorsal post- 
cleithrum is the larger, the ventral merely a narrow splint of bone. 

The pectoral fin is composed of 14 rays, the first being the longest. All the rays 
are segmented and most of them, if not all, appear to be branched. 

Pelvic girdle and fin. The triangular pelvic bones are sutured together in the mid- 
ventral line. Each is thickened posteriorly to form a prominent condyle supported 
by a thickened, lateral ridge. The pelvic fin has 9 rays, all of which are segmented, 
but several appear to have been unbranched. At the base of the fin is a small 
asymmetrically placed pelvic splint bone. 

Vertebral column. The vertebral column is shown in the restoration, Text-figure 
71. It is composed of 30 vertebrae, of which 16 are caudal. The centra are 
slightly longer than deep and mesially constricted. The neural arches are not 
fused to the centra anteriorly, although the more posterior neural arches have a 
more intimate connection. The neural spines originate from the anterior end of 
each neural arch. The precaudal vertebrae bear very small transverse processes 
which elongate on the last three or four precaudal vertebrae. Long, curved, 
flattened pleural ribs articulate with the small transverse processes. The haemal 
arches are extended into prominent, curved haemal spines. Zygapophyses are 
present on both the dorsal and ventral surfaces of the vertebrae, but are most 
pronounced in the ventral caudal region. Epineurals and epipleurals are present 
on the first 20 vertebrae. 

Median fins and tail. The median fins are shown in the restoration, Text-figure 
71. The dorsal fin consists of 11 rays and is situated in the second quarter of the 
back. The second or third ray is the longest and much in excess of the basal length 
of the fin. The first ray is short and unbranched, the second is longer and also 
unbranched, although both are segmented. The subsequent rays are branched and 
segmented. The radials are long, flattened and unexpanded except in their articula- 
tory head regions. Between the occiput and the origin of the dorsal fin, four accessory 
radials are present. The most anterior accessory radial is associated with the 
neural spine of the first vertebra. This association between accessory radial and 
neural spine may only have been ligamentous. 

The anal fin is positioned behind the level of the rear of the dorsal fin and mid- 
way between the pelvics and the caudal. Nine rays are present in the anal fin ; the 
first is short and unbranched although segmented, the second is long and unbranched, 
and the remainder of the rays are branched and segmented. 



ESPECIALLY MYCTOPHOIDS 159 

The caudal skeleton (Text-fig. 72) consists of seven vertebrae. Four free preural 
vertebrae are present. Preural vertebra 1 is fused with ural vertebra 1, and the 
second ural vertebra is present as a terminal half-centrum. Preural vertebrae 3, 
4 and 5 bear slightly elongated neural and haemal spines which are somewhat 
expanded. Preural vertebra 2 (the first free preural vertebra) has an expanded 
haemal spine and no neural spine, but the neural arch is enlarged. The fused 
preural and ural vertebra bears the parhypural and the first two hypurals ventrally. 
Hypural 1 is larger than hypural 2, the latter not extending to the mid-line of the 
caudal fin. The second ural vertebra supports the upper four hypurals. The lower 



n.a.pu2 




Fig. 72. Sardinioides minimus (Agassiz). Caudal fin skeleton in lateral view. 

edge of hypural 3 does not reach the mid-line of the caudal fin, and thus there is a 
small gap left between the upper and lower hypural elements. The fused preural 
and ural centrum bears an enlarged, anteriorly expanded stegural. This is inclined 
postero-dorsally and a second uroneural is present behind it. Three epurals 
associate with the dorsal edge of the anterior expansion of the stegural. 

The caudal fin consists of 19 principal rays of which 17 are branched. Ten 
accessory rays occur epaxially and 10 hypaxially. In front of the anteriormost 
accessory ray, both dorsally and ventrally, is an enlarged fulcral scale. 

Scales. The entire body is covered with cycloid scales which extend on to the 
opercular bones. Scales also cover the cheek region, the posterior end of the 
maxilla, the articular region of the mandible, and extend on to the dorsal surface of 
the skull as far as the centre of the orbit. In the smaller specimens the scales are 
strictly cycloid, but in larger specimens the scales may become very feebly ctenoid. 



160 UPPER CRETACEOUS TELEOSTS 

However, even on the larger specimens, not all the scales are ctenoid, those scales 
on or near the head are cycloid, the ctenoid scales being confined to the mid-body 
region. 

Order CTENOTHRISSIFORMES 

Family AULOLEPIDAE Patterson, 1964 

Diagnosis. See Patterson (1964 : 247), except that post-temporal fossa is occa- 
sionally roofed. 

Genus PATTERSONICHTHYS nov. 

Diagnosis. Aulolepidae in which the post-temporal fossa is roofed. Length of 
the head greatly exceeds the depth of the trunk. Operculum devoid of scales. 
Vertebrae 30 in number of which 17 are caudal. Dorsal fin with 13 or 14 rays ; 
anal fin behind rear of dorsal with 10 rays. 

Type species. Pattersonichthys delicatus. 

Remarks. The genus Pattersonichthys is erected to contain several small speci- 
mens from Hajula, Lebanon. The generic name is dedicated to Dr. Colin Patterson 
of the B.M.N.H. 

Pattersonichthys delicatus sp. nov. 

(Text-figs. 73-77) 
Diagnosis. As for genus, only species. 

Holotype. B.M.N.H. specimen number P. 9976, a complete but flattened fish, 
from the Middle Cenomanian, Hajula, Lebanon. 

Material. The holotype and two further specimens, P. 9977a and P.99776, also 
from the Middle Cenomanian, Hajula. All three specimens were prepared in 
acetic acid by the transfer method. 

Description. Neurocranium. The neurocranium appears relatively shallow, 
and is deepest in the occipital region. The parasphenoid, visible in the base of the 
orbit, is straight. The frontals are flat, although the lateral edges curve down to 
form the dorsal margin of the orbits. The frontal tapers anteriorly to insert on to 
the dorsal surface of the mesethmoid and is unornamented. In the posterior region 
of the cranial cavity an incompletely roofed post-temporal fossa is present. The 
roof of the fossa has a curved indentation posteriorly. The frontal extends laterally 
above the dorsal surface of the sphenotic at the hind end of the orbit, and medially to 
cover the anterior half of the cranial cavity. Posteriorly the frontals contact the 
parietals and the pterotics. The pterotic makes up most of the postero-lateral 
edge of the neurocranium, meeting the sphenotic anteriorly and the frontal and 
parietal medially. The lateral faces of the pterotic and sphenotic are deeply 
excavated to form the dilatator fossa. The upper border of the fossa is delimited 
by a low crest on the pterotic passing backwards from beneath the frontal. The 



ESPECIALLY MYCTOPHOIDS 



161 



hyomandibular facet is divided into two regions, an anterior cup-shaped concavity 
in the sphenotic and prootic, and a posterior elongated groove on the ventro-lateral 
surface of the pterotic. 

The large parietals meet in the mid-line. The parietal contacts the frontal 
anteriorly and the pterotic laterally and enters into the composition of the roof of 
the post-temporal fossa. The remainder of the neurocranium is poorly shown. 
The epiotic appears to project posteriorly and the dorsal limb of the post-temporal 
bone is associated with its dorsal surface. In the region of the occipital condyle the 
exoccipital and basioccipital are thickened and meet in a prominent suture. Below 
the condylar region the parasphenoid is curved upwards slightly. In the base of 
the orbit the parasphenoid is expanded laterally into ventrally inclined flanges. 

The mesethmoid is prominent but appears to have been only superficially ossified. 
It bears a pair of lateral wings inclined antero-ventrally, which are underlain by 
lateral extensions of the vomer. There is no connection laterally between the two 
bones. The palatine, antero-medially, is closely associated with the lateral edge of 
the vomer and also possibly the mesethmoid. The lateral ethmoids are not well 
ossified and ventrally contact the palatines. The lachrymal lies on the ventro- 
lateral region of the lateral ethmoid. 

The articular head of the maxilla appears to have rested on the dorsal surface of 
the lateral part of the vomer and would thus also have contacted the ventral surface 
of the lateral wing of the mesethmoid. The posterior process or shaft of the vomer 



os.p.pm 



mx.p.p 



/.pel 




smx2 



Fig. 73. Pattersonichthys delicatus gen. et sp. nov. Restoration of the skull of the holotype 
(B.M.N.H. number P.9976) in lateral view. 



162 



UPPER CRETACEOUS TELEOSTS 



extends back beneath the anterior parasphenoid region. Whether the vomer is 
toothed could not be seen. 

Within the postero-ventral region of the orbit there is a long basisphenoid pedicel, 
but no evidence of an orbitosphenoid could be seen. At the anterior end of the 
snout a small nasal bone lies lateral to the anterior region of the frontal and extends 
on to the dorsal surface of the mesethmoid. 

Infraorbital bones. A circumorbital ring of bones is present although they are 
poorly preserved and fragmentary. There are five or six bones in the series, 
including the lachrymal. The bones are narrow and roughly equal in size except 
for the lachrymal which is larger. The bones are grooved laterally for the infra- 
orbital sensory canal. The last infraorbital articulates with the sphenotic. Dorsal 
to the lachrymal is a further bone which probably represents an antorbital. The 
lachrymal and the first infraorbital show signs of having possessed a slight subocular 
shelf. 

Hyopalatine bones. The hyomandibular is slender and vertical with a single 
articulatory head. The opercular process is small, in the upper third of the bone, 
and covered laterally by the preoperculum. Anteriorly the hyomandibular is thin 
but its posterior edge is considerably thickened. Just within the upper half of the 
posterior edge is a short process projecting postero-laterally from the crest which 
runs along the posterior edge of the bone. The hyomandibular tapers ventrally and 
a stout symplectic bone connected it with the quadrate. The symplectic is attached 




smxl 



smx2 



Fig. 74. Patter sonichthys delicatus gen. et sp. nov. Restoration of the skull in lateral view. 

From B.M.N.H. number Y'.ggj-ja. 



ESPECIALLY MYCTOPHOIDS 163 

to the medial face of the quadrate and inclines forwards at an angle to the hyo- 
mandibular. The quadrate is large and triangular with the anterior edge con- 
siderably longer than the posterior. The condyle is thickened and mesially con- 
stricted. The posterior edge of the quadrate is thicker where the preoperculum 
rests against it. The postero-dorsal edge of the quadrate is deeply excavated at 
the upper limit of the symplectic groove. The quadrate dorsally contacts the lower 
edge of the metapterygoid, and anteriorly the ectopterygoid. The remainder of the 
palate is not preserved, except anteriorly where the maxillary process of the palatine 
is visible. This process is in the form of a funnel-shaped protuberance from the 
dorso-lateral part of the palatine. 

Dermal upper jaw. The dermal upper jaw is shown in lateral view in Text- 
figures 73 and 74. It is long and extensively toothed. The premaxilla has a short, 
broad, ill-defined ascending process. Laterally the ascending process has a small 
articulatory process for a facet on the leading edge of the maxillary head. The 
posterior process of the premaxilla is prominent with a broad oral border covered 
by small recurved teeth. The premaxilla forms the anterior third of the oral border 
of the upper jaw. The external face of the premaxilla is unornamented. 

The maxilla is also unornamented. Anteriorly the maxilla curves medially and 
has a large rounded dorso-medial depression, bounded posteriorly by a prominent 
flange. The maxillary process of the palatine is contained within this facet. 
Antero-medial to the facet the head of the maxilla is rounded. The maxillary head 
appears to be associated with the dorsal surface of the vomer and the undersurface 
of the mesethmoid. The anterior face of the maxilla, in front of the palatine facet, 
is excavated for the reception of the articular process of the premaxilla. The maxilla 
extends posteriorly above the premaxilla as a stout bar of bone. Behind the pre- 
maxilla the maxilla deepens, and the tooth-bearing margin is thinner and flatter 
although the stout part of the maxilla is continued back for some way above the 
dentigerous region. The toothed margin is a deep, curved flange ; the teeth are 
numerous anteriorly, but reduced in number towards the posterior end. 

Two large supramaxillae are attached to the dorsal edge of the maxilla along 
most of its length. The anterior supramaxilla is a roughly oval slip of bone pro- 
longed somewhat anteriorly. The posterior supramaxilla is larger and has a dorsal 
process extending anteriorly along the upper edge of the anterior supramaxilla. A 
ridge on this antero-dorsal process continues back on to the lateral face of the 
supramaxilla. Both bones are unornamented. 

Mandible. The mandible is shown in lateral view in Text-figures 73 and 74. It 
is quite long and deep, with a pronounced coronoid process. The symphysis is 
blunt and not deepened, although there seems to be a slight ventral prolongation. 
The oral border of the dentary bears a group of small, clustered, recurved teeth. 
The mandibular sensory canal, in the dentary, ran within the bone, opening to 
the exterior through three or four prominent pores. The tube opens into a groove 
towards the posterior end of the dentary. The articular forms the posterior part of 
the mandible and the articular facet is shallow and transversely orientated. A small 



i6 4 UPPER CRETACEOUS TELEOSTS 

insignificant retroarticular process occurs behind the facet and a groove on the lateral 
face of this process extends below the articular facet and passes forwards on the 
ventro-lateral surface of the articular. This groove carried the sensory canal. The 
posterior edge of the articular rises steeply from the facet to form the posterior part 
of the coronoid process. The angular is present as a small slip of bone applied to the 
postero-ventral region of the articular below the facet. The lateral face of the 
mandible is smooth and unornamented. 

Opercular bones. The opercular bones are shown in lateral view in Text-figures 
73 and 74. The preoperculum inclines forwards ventrally and dorsally ends just 
behind and below the head of the hyomandibular. The anterior edge of the pre- 
operculum lies against the posterior face of the hyomandibular crest. Ventral to the 
hyomandibular, the anterior edge of the preoperculum is expanded into a small 
convexity. Below this slight expansion the preoperculum lies against the rear edge 
of the quadrate. The preopercular sensory canal ran in a deep groove, open poster- 
iorly. A thin flange of bone overhangs this groove, neither the flange nor any other 
part of the bone is ornamented. 

The operculum is deeper than long with a straight anterior edge which terminates 
in an acute point antero-ventrally. The ventral edge is obliquely inclined, whereas 
the rest of the margin is smoothly rounded. The opercular facet occurs just within 
the upper half of the bone and is supported by a horizontal strengthening ridge 
which passes across the inner face of the operculum. The lateral face of the oper- 
culum is unornamented. 

The suboperculum is large with its dorsal edge lying medial to the operculum. It 
extends postero-dorsally for almost half of the depth of the operculum. The 
antero-dorsal corner of the suboperculum is produced into a small pointed process. 
The suboperculum is thin and delicate, except for a thickened anterior edge. 

The interoperculum is a roughly triangular sheet of bone with a straight posterior 
edge which abuts against the anterior edge of the suboperculum. The inter- 
operculum is thin and unornamented. 

Hyoid arch and branchiostegal rays. The ceratohyal is large and rectangular in 
shape, with a deep lateral groove running along its length. The groove indicates 
the course of the hyoidean artery and is interrupted by an elongate fenestra approxi- 
mately mid-way along the ceratohyal. There are two hypohyals. The exact 
number of branchiostegals is uncertain but they are arranged in two distinct groups. 
The ceratohyal supports several small, thin branchiostegal rays, possibly as many as 
five. The second group of rays arises from the epihyal and one from its junction 
with the ceratohyal. The branchiostegal rays on the epihyal are broad, long, 
flattened and four in number. 

Pectoral girdle and fin. The pectoral girdle is shown in lateral view in Text-figure 
75. The post-temporal has a long dorsal limb articulating with the epiotic and 
a shorter, thinner ventral limb passing antero-ventrally. The junction of the 
two limbs is somewhat flattened and expanded. No trace of a supratemporal 
could be found in any of the specimens. 



ESPECIALLY MYCTOPHOIDS 



165 



The supracleithrum is stout and narrow and ventrally it articulates with the 
lateral face of the cleithrum, and is confined within a vertical groove. The cleithrum 
is a prominent sigmoid bone expanded lateral to the attachment of the endoskeletal 
girdle. The ventral part of the cleithrum curves forwards, and its anterior edge is 
bent medially. 

The endoskeletal girdle is indistinct, although the thickened saddle-shaped facet 
on the postero-dorsal region of the scapula, with which the first fin ray articulated, 
is clearly defined. The coracoid has a long slender anterior process which curves 
forwards to the anteriormost tip of the cleithrum with which it connects. The 
ventral edge of the coracoid projects posteriorly as a short pointed spine below the 
fin insertion. 



d.pcl 



v. pel 




2mm 



Fig. 75. Patter sonichthys delicatus gen. et sp. nov. Pectoral girdle of the right side in 
lateral view. From B.M.N.H. number P.9976. 



The four radials articulate with the rear edge of the scapula, and the junction 
between the scapula and coracoid. Two prominent postcleithra are present ; 
the larger dorsal one articulates with the medial face of the posterior expansion of 
the cleithrum, the ventral postcleithrum is smaller and joined to the upper post- 
cleithrum dorsally and the pelvic bone ventrally. 

The pectoral fin consists of 13 rays of which the first is the longest, the remainder 
decreasing in size. All of the rays are segmented distally, and most of them appear 
to be branched. 

Pelvic girdle and fin. The pelvic bones are flattened, widest posteriorly and 
tapering to a point anteriorly. They contact each other in the mid-line in a long 



166 UPPER CRETACEOUS TELEOSTS 

suture. Although thin the pelvic bones are thickened somewhat posteriorly to form 
an articulatory condyle. A crest passes anteriorly along the lateral edge of the bone. 
The main body of the bone lies horizontally whereas the condyle is inclined obliquely. 
Medial to the condyle, the ventral postcleithrum is firmly attached. The anterior 
ends of the pelvic bones lie between the anterior coracoid processes. 

The pelvic fin is long, prominent and composed of 8 rays. All the rays are seg- 
mented distally, and except for the first one or two are branched. The fin has a 
basal, asymmetrical, pelvic splint bone. 

Vertebral column. The vertebral column is shown in the restoration, Text-figure 
76. It consists of 30 vertebrae, of which 17 are caudal. The centra are longer 
than deep, and mesially constricted. The neural arches are relatively stout but 
the spines are delicate. Each neural arch is confined to the anterior half of the 
corresponding centrum. The neural arches on the precaudal vertebrae are only 
loosely connected to the centra, but caudally the contact is much closer. Transverse 
processes only become evident on the posteriormost two or three precaudal vertebrae, 
and the pleural ribs on the remaining vertebrae articulate directly with the centra. 
Several of the most anterior precaudal vertebrae do not appear to possess pleural 
ribs. On the first caudal vertebra the haemal arch is not drawn out into a haemal 
spine, but on all the subsequent caudal vertebrae haemal spines are present. Zyga- 
pophyses are in evidence on the ventral regions of both precaudal and caudal 
vertebrae, whereas on the dorsal region of the vertebral column the zygapophyses 
are non-existent. Epineurals and epipleurals are present on the first 22 vertebrae. 

Median fins and tail. The median fins are shown in the restoration, Text-figure 
76. The dorsal fin is prominent and occurs in the second quarter of the back, 
consisting of 13 or 14 rays. The fifth ray is the longest and all the subsequent rays 
are segmented. The first four rays decrease in size anteriorly. The third and 
fourth rays are unbranched but segmented, but the first and second rays are neither 
branched nor segmented. The radials are long and narrow and unexpanded. The 
medial radials are visible between the bases of adjacent fin rays, as elongate con- 
stricted cylinders of bone inclined obliquely. In the region between the origin of 
the dorsal fin and the occiput three flattened accessory radials occur. The most 
anterior one is associated with the expanded neural spine of the first vertebra. 

The anal fin is composed of 10 rays and arises just behind the rear of the dorsal fin. 
The third ray is the longest. It is greatly extended and both segmented and distally 
branched. The subsequent rays are all branched and decrease in size posteriorly. 
The second ray is shorter and segmented, the first is the shortest and spinous in form. 

The caudal skeleton (Text-fig. 77) is composed of several vertebrae. Four 
preural vertebrae are present as separate units, the neural and haemal spines of 
these supporting the accessory rays of the caudal fin. The first preural vertebra is 
fused with the first ural vertebra. A free second ural vertebra is present as a small 
terminal half-centrum posteriorly. The haemal spines of the preural vertebrae are 
elongate, that of preural centrum 2 (the first free preural vertebra) is slightly 
flattened and expanded. The fused preural and ural vertebrae support the par- 
hypural (the haemal spine of the first preural vertebra) and two hypurals. Postero- 



ESPECIALLY MYCTOPHOIDS 



167 







« 



> 

o 

0<3 

-M g 

d 

to 
S 



13 



£ 



£ 



f* 



1 68 



UPPER CRETACEOUS TELEOSTS 



dorsally the parhypural contacts the centrum by a narrow connection, and anteriorly 
overlaps the haemal spine of the second preural vertebra. The parhypural is long, 
narrow and flattened. The two hypural bones support the lower caudal lobe, 
hypural i is the larger. Hypural 2 is narrow and does not extend dorsally to the 
mid-line of the fin. The second ural vertebra supports the four hypurals of the 
upper caudal lobe. These hypurals have partially united to form a hypural plate 
which does not extend ventrally to the mid-line of the fin. There is thus a distinct 
gap in the mid-line of the caudal skeleton between the upper and lower hypural 
elements. Long, narrow neural spines occur on preural vertebrae 3, 4 and 5. The 
neural spine of preural vertebra 2 is missing and the neural arch is considerably 
expanded. Two uroneurals are associated with the dorsal surface of the terminal 
vertebrae. The stegural is formed from the anteriorly expanded first uroneural, and 
it is drawn out postero-dorsally. At least two, and possibly three, epurals are 
associated with the dorsal edge of the stegural. 

The caudal fin has 19 principal rays, 17 of which are clearly branched. Nine 
accessory rays are present in front of the upper caudal lobe, and 9 below. The 
anteriormost accessory ray both above and below is slightly more flattened and more 
horizontally placed. This anterior ray may represent the remains of a fulcral scale. 

Scales. The entire body is covered with small cycloid scales which do not appear 
to show any ctenoid tendencies. Scales are also present on the cheek region but not 
on the opercular bones. 



n.a.pu2 




Fig. 77. Patter sonichthys delicatus gen. et sp. nov. Caudal fin skeleton in lateral view. 



ESPECIALLY M YCTOPHOIDS 169 

III. DISCUSSION 

This section is devoted to a discussion of the new families and suborders erected 
in the systematic descriptions (Part II). Each group is considered separately. 



Order SALMONIFORMES 

(A) Suborder ICHTHYOTRINGOIDEI 

A new suborder has been erected to contain the three genera Ichthyotringa, 
Apateodus and Apateopholis as set out in the systematic descriptions. The suborder 
has been divided into two separate but closely related families, the Ichthyotringidae 
and the new family Apateopholidae. Despite the similarities between the two 
(compare Text-figs. 3 and 15), the Apateopholidae is somewhat more advanced. 
The advanced characters form the basis of the familial division : — 



ICHTHYOTRINGIDAE APATEOPHOLIDAE 

Complete roof to the post-temporal fossa. Post-temporal fossa unroofed. 

Supratemporal present but reduced. Supratemporal absent. 

Orbitosphenoid present. Orbitosphenoid absent. 

Basisphenoid present. Basisphenoid absent. 

Posterior infraorbitals expanded. Infraorbitals much reduced in extent. 

Jaw suspension vertical. Jaw suspension forwardly inclined. 

No cranial deepening. Cranium deepened. 

Complete covering of cycloid scales. Body naked except for a row of lateral line 

scales. 



Apart from these features the two families are linked by several primitive features 
of the body and tail, and by several advanced characteristics associated with the 
production of an elongate rostral region. 

The rostrum is composed of premaxillae, maxillae, palatines, vomer and meseth- 
moid. In both families simplification has occurred by the loss of the nasals (or their 
combination with the mesethmoid) and the loss of both antorbitals and supraorbitals. 
Teeth are borne on the palatines and the premaxilla but not on the vomer. This 
rostral composition is similar to that seen in the dercetids (pp. 50-71) and certain 
of the ' alepisauroid myctophiforms ' (Parr, 1929; Marshall, 1955). 

The general form of the body is the same in both families and is short, relative 
to the rostral elongation, with approximately 40 vertebrae. The anterior neural 
spines are composed of separate lateral elements ; intermuscular bones occur 
along most of the column ; the dorsal fin is in the mid-region of the back ; the anal 
fin is remote ; the pelvics are abdominal, lying below the dorsal fin ; the pectoral 
fins are set low on the flanks and consist of a large number of rays. In addition to 
these features the caudal skeleton contains two free ural vertebrae, thus corres- 



170 UPPER CRETACEOUS TELEOSTS 

ponding to the condition seen in Elops (Nybelin, 1963) and some basal salmoniforms 
such as Salmo (Norden, 1961). 

The Ichthyotringidae possess a complete covering of thin cycloid scales, whereas 
in the Apateopholidae the scales are confined to the lateral line only. This latter 
condition may indicate that some degree of stomach distensibility was possessed 
by Apateopholis. 

The genus Apatcodus is placed as an addendum (incertae sedis) to the family 
Ichthyotringidae. It is unfortunate that the genus is only known from skull 
material. The head appears to agree in certain features with that of Ichthyotringa, 
and is closer to this genus than to any other Cretaceous form so far described. 
Features shared by the two genera are as follows : 

1. Parietals meet in the mid-line of the skull-roof. 

2. Post-temporal fossa completely roofed. 

3. Mesethmoid simple and pointed anteriorly. 

4. Orbitosphenoid present. 

5. Basisphenoid present. 

6. Posterior infraorbitals greatly expanded, covering the hyomandibular 

and extending back to the preoperculum. 

7. Rostral region from the same components, although in Apateodus it is con- 

siderably shorter. 

The three genera placed in the suborder Ichthyotringoidei have not previously 
been associated together. The family Ichthyotringidae was first proposed by 
Jordan (1905) to include Cope's (1878) genus Ichthyotringa. Woodward (1901 : 
265) placed the genus Ichthyotringa (Rhinellus) within his family Scopelidae among 
the Isospondyli. In the same work (1901 : 258) Woodward associated Apateodus 
with Ichthyotringa. Later, (1902 : 38) using more complete material, he put the 
genus Apateodus in the Enchodontidae. Regan (191 1 : 120) considered Apateodus 
to be close to the genus Alepisaurus and concurred with Woodward's (1902) view 
that Apateodus belonged with the enchodonts. Kruizinga (1924) in his more 
extensive study of the head of Apateodus retained the genus within the enchodonts 
but related the family to the stomiatoids. Gregory (1933 : 204) mentioned 
Apateodus in connection with the enchodonts contained in his division Mesichthyes. 

Jordan (1923 : 155) later placed the genus Ichthyotringa in the Myctophidae 
among the Iniomi, but Berg (1940 : 438) retained the familial name and it was 
contained in his order Scopeliformes. Bertin and Arambourg (1958 : 2272) placed 
the genus Ichthyotringa within the family Myctophidae of the suborder Myctophoidei. 

Some mention of the historical aspect of the genus Apateopholis has already been 
given in the systematic description. The genus has not been considered by any 
author since Woodward (1901 : 232) placed it in the family Enchodontidae. 

Neither Apateodus nor Apateopholis can be considered to be enchodonts (as con- 
stituted in this work) due primarily to the difference in palatine structure. The 
Enchodontoidei are characterized by the presence of a single, terminal tooth on the 
palatine. Many other structures are also at variance with the enchodontoids, for 
example the caudal skeleton and premaxillary arrangement. 



ESPECIALLY MYCTOPHOIDS 



171 



Likewise Ichthyotringa cannot be retained in the myctophiform group (in which it 
has often been placed) because of the following characters : 

1. Maxilla enters the gape behind the premaxilla and is toothed. 

2. Premaxilla simple, with no ascending process. 

3. Posterior infraorbitals expanded. 

4. Caudal skeleton with two free ural vertebrae. 

These four characters provide ample justification for separating the Ichthyo- 
tringidae from the Myctophiformes. Additional substantiation for the separation 
is provided by the time of occurrence of the two groups in the fossil record. 
Apateodus first appears in the Albian (Lower Cretaceous) ; Ichthyotringa in 
both the Lower Cenomanian {Ichthyotringa africana, Arambourg 1954) and the 
Middle Cenomanian (Ichthyotringa delicata, Hay 1903) ; whilst Apateopholis, the 
most specialized of the ichthyotringoids, is also from the Middle Cenomanian. 
Thus the Ichthyotringoidei had not only arisen but had radiated to produce highly 
specialized members by the beginning of the Upper Cretaceous. At an equivalent 
time (i.e. Middle Cenomanian) the myctophiform lineage is only represented by basal, 
generalized forms such as Sardinioides. If the Ichthyotringoidei and the Myctophi- 
formes are related then the myctophiforms would presumably have arisen from the 
ichthyotringoids. This view is clearly untenable since the Ichthyotringoidei 
possess several specialized characters which are more advanced than anything seen 
in the basal myctophiformes. These features include : 

1. The long rostrum with its entailed loss of bony elements. 

2. Absence of fulcral scales at the base of the tail. 

3. Absence of a pelvic splint bone. 

4. Absence of accessory radials in front of the dorsal fin radials. 

These four factors are, however, encountered among some of the more advanced 
' alepisauroid myctophiforms' (Marshall, 1955; Gosline, Marshall & Mead, 1966). 
For example rostral lengthening and the forward migration of the jaw articulation 
occur in Lestidium and Anotopterus (Text-fig. 78) and the similarity in jaw structure 
between Anotopterus and Ichthyotringa has been pointed out by Arambourg (1954 : 




pop q drt den 

Fig. 78. Anotopterus pharao. Skull in lateral view. 



•72 



as.p.pm 



UPPER CRETACEOUS TELEOSTS 
fr 



hm 




p.p.hm 



pop 



Fig. 79. Alepisaurus ferox. Skull in lateral view. 



124). Regan (191 1 : 120) has noticed the similarity between Apateodus (Text- 
fig. 11) and Alepisaurus (Text-fig. 79) in both of which no great rostral elongation 
has occurred. The composition of the snout in these two genera is also similar 
with an extended mesethmoid, palatine and vomer with the palatine supporting 
several enlarged, laterally compressed teeth. One difference is that in Apateodus 
both palatine and ectopterygoid bear teeth, whereas in Alepisaurus all of the 
teeth are borne on the palatine. The dermal upper jaws are superficially similar 
but Apateodus is more primitive in that the premaxilla does not have an ascending 
process, and the maxilla is a simple strut. In Alepisaurus the premaxilla and 
maxilla have been derived by simplification from a much more complex pattern. 
Both genera have lost the antorbitals and nasals due to the elongation of the snout. 
The caudal skeleton has already been mentioned as being in a condition com- 
parable to the elopoids and salmonoids in respect of the number of separate ural 
vertebrae. The caudal skeleton of Ichthyotringa furcata (Text-fig. 4) shows a neural 
arch without a neural spine on the first preural vertebra which is like that seen in Elops 
(Text-fig. 80) and Sedenhorstia (Goody, 1969, fig. 2). In the Salmoniformes the 
neural arch of this vertebra is incorporated with the first uroneural (and probably the 
neural arch of ural vertebra one) to produce the stegural (Text-fig. 81). In Ichthyo- 
tringa the neural arch of preural vertebra one is distinct and the anterior uroneural is 
expanded anteriorly. This expansion probably represents the neural arch of ural 
vertebra one and thus an earlier stage in the production of a stegural, with the neural 
arch of preural vertebra one not having yet been incorporated. In Elops there is also 



ESPECIALLY MYCTOPHOIDS 



173 



a small anterior expansion in advance of the first uroneural, which may well represent 
the neural arch of ural vertebra one. If this is the case then the forerunner of a 
stegural is present in the elopoids. In the Salmoniformes the epurals are associated 
with the dorsal edge of the stegural. In Elops and Ichthyotringa the epurals contact 
the dorsal edge of the anterior expansion of uroneural one. This possibly affords 
further evidence that the anterior expansion is the forerunner of a true stegural. 



n.a.ul 



n. a. pul 



r.n.s.pu2 



pu4 




5mm 



Fig. 80. Elops sanrns. Caudal fin skeleton in lateral view. 



Preural vertebra two in Ichthyotringa has a complete neural spine. In Elops this 
neural spine is reduced (Text-fig. 80), as it is in certain salmonoids (Norden, 1961; 
pi. 14, fig. G, Coregonus artedii). In other salmonoids the neural spine is complete 
(Text-fig. 81). In those forms in which a reduced neural spine occurs then three 
separate epurals are normally seen (the primitive teleostean number according 
to Patterson, 1968a : 220). Ichthyotringa, with its complete neural spine on preural 
vertebra two, has only two epurals. It would seem possible, then, that the anterior- 
most epural, when three are present, comes from the neural spine of preural vertebra 
two, the remaining two epurals being the neural spines of preural vertebra one and 
ural vertebra one. Patterson (1968a : 221) has drawn attention to the fact that the 
foremost epural can " fuse with a neural arch so that it is indistinguishable from a 
neural spine ". It is equally possible that this is what has occurred in Ichthyo- 
tringa. 



'74 



UPPER CRETACEOUS TELEOSTS 



Ichthyotringa agrees with Elops but not with the salmonoids in the possession of 
two long, narrow, rod-like uroneurals. Elops has a third smaller uroneural postero- 
dorsally, but this is not observed in Ichthyotringa, although present in Saltno 
(Norden, 1961). 




Fig. 81. Salmo gairdneri. Caudal fin skeleton in lateral view. After Norden (1961). 



Six hypurals are evident in Ichthyotringa and in this respect it is comparable to the 
salmonoids. The elopoids have seven hypurals, the seventh very small and 
practically hidden beneath the third uroneural. The arrangement and size of the 
hypurals is of characteristic form in Ichthyotringa. The two ural vertebrae are much 
reduced in length and strongly upturned, much more so than in either Elops or 
Salmo. In both of these latter genera the hypurals decrease in length from hypural 
one to hypural six (or seven). In Ichthyotringa the hypurals remain quite long and 
there is no equivalent to Monod's (1967) " hypurale minimum ", the small isolated 
last hypural. All of the hypurals in the upper caudal lobe of Ichthyotringa appear 
to attach to the small second ural vertebra. 

All of the elopoids and the salmonoids possess 17 branched caudal rays, as does 
Ichthyotringa. In the Ichthyotringoidei and the Salmonoidei no fulcral scales occur, 
but these are present in the elopoids. 

A feature noticed in the basal Salmoniformes is the tendency for the neural and 
haemal spines in the posterior caudal region to be considerably laterally compressed 
and expanded (Gosline, i960 ; Norden, 1961 ; Vladykov, 1962). In the elopoids as 
well as in Ichthyotringa the posterior neural and haemal spines are long and narrow 
and drawn out in a posterior direction with no compression or expansion. 

On tail structure the Ichthyotringoidei show marked similarities to the Elopi- 
formes, and in other characters also resemble the elopoids, for example : 



ESPECIALLY MYCTOPHOIDS 



175 



Roofed post-temporal fossa. 

Orbitosphenoid and basisphenoid present. 

Posterior infraorbitals expanded. 

No ascending process on the premaxilla. 

Mesocoracoid arch (Apateodus). 

Abdominal pelvic fins. 

Supraorbital bone (Apateodus). 



These characters are all, however, merely indicative of a primitive grade of organi- 
sation, and could equally ally the Ichthyotringoidei with the salmonoids. Certain 
features, however, definitely separate the ichthyotringoids from the elopoids : 

1. No rostral bones, and no ethmoidal commissure. 

2. No antorbital. 

3. No nasal. 

4. No supramaxillae. 

5. No gular plate. 

6. Development of a trigemino-facialis chamber in the prootic (Apateodus) 

as opposed to the separate intraosseus passages seen in the elopoids 
(Greenwood, et al., 1966). 

7. No teeth on the parasphenoid. 

These features are sufficient clearly to separate the ichthyotringoids from the 
elopoids and are features which are present in the more progressive salmoniforms. 
Greenwood, et al. (1966 : 350), in the erection of their new superorder Protacantho- 
pterygii, suggested that the salmonoids constitute the principal and basal element of 
this group, possessing none of the limitations of the elopoids, and the salmonoids 
would appear to have the requirements to give rise to the major radiations within 
the Protacanthopterygii and Acanthopterygii. The salmonoids possess the following 
basal osteological characters : parietals in contact in the mid-line ; orbitosphenoid 
and basisphenoid ; supraorbital and antorbital ; expanded posterior infraorbitals ; 
maxilla large and toothed ; incipient ascending process of premaxilla ; two large 
supramaxillae ; teeth on vomer, palatine, endopterygoid and ectopterygoid ; 
numerous branchiostegal rays ; mesocoracoid arch ; three postcleithra ; as many 
as 18 pectoral rays ; anterior neural spines consisting of separate lateral elements ; 
numerous supraneurals ; numerous epineurals ; transverse processes not fused 
to centra ; pelvic fin with n rays ; pelvic splint bone ; no fin spines ; two free 
ural vertebrae ; stegural with one or two further uroneurals ; 17 branched caudal 
rays ; six hypurals ; cycloid scales. 

The Ichthyotringoidei agree with the majority of these characters, differing in the 
features related to rostral elongation, and the absence of a stegural, accessory 
radials and a pelvic splint bone. In conclusion, then, the ichthyotringoids represent 
a small radiation commencing in the Lower Cretaceous. Apateodus is the least 
specialized genus and continued little changed throughout the Upper Cretaceous 
(Gault to Maastrichtian). It was a generalized predator which may possibly have 
given rise to the other more advanced ichthyotringoids. The elongate rostral region 



176 UPPER CRETACEOUS TELEOSTS 

does not appear to represent as efficient a modification for predatory feeding as does 
the more mobile and protrusile acanthopterygian jaw. The rise of the acanthoptery- 
gians probably accounted for the disappearance of the ichthyotringoids in the Upper 
Cretaceous. The more generalized Apateodus, however, existed until the extreme 
end of the Cretaceous and outlasted its more specialized descendants. The only 
recent teleosts which possess a long snout and an unelongated body seem to be the 
Xiphiidae, possibly strengthening the argument that this rostral elongation 
represents a less successful way of life. 

The Ichthyotringoidei must have arisen as a very early offshoot from a basal 
salmoniform stock, a stock not too far removed from the elopiforms. 



(B) Suborder CIMOLICHTHYOIDEI 

This new suborder has been erected to contain several quite closely related genera, 
all of which are difficult to link with any Recent forms or with any so far described 
fossil forms. The two families Cimolichthyidae and Dercetidae show great similari- 
ties in overall skull and body structure, in particular the rostral region and the 
reduction of the body scaling to two or three isolated scute rows on each flank (Text- 
fig. 82). Both families occur in the Cenomanian and extend through much of the 
Upper Cretaceous. Many incomplete and fragmentary specimens of large Dercetis 
species from the English Chalk (Woodward, 1903), e.g. Dercetis maximus and Dercetis 
latiscutatus, are closely related to Cimolichthys levesiensis from the same horizons. 
These large species of Dercetis have not been considered in the systematic account 
since they are represented by very imperfect material. 

The structurally interesting features shown by the cimolichthyids and dercetids 
are mainly concerned with the extension of the head or body or both. These 
elongations of the head and body have occurred to a greater or lesser extent in all of 
the genera. Cimolichthys shows slight snout and body elongation ; Dercetis has a 
variably extended snout with a long body (Text-figs. 23-26) ; Rhynchodercetis has 
an enormously elongated snout and a very long body (Text-figs. 28 and 29), and 
Pelargorhynchus has some snout elongation but much body lengthening. 

Snout elongation is always brought about by the extension of the preorbital region 
and includes the vomer, mesethmoid, palatines, premaxillae, maxillae and also the 
frontals to a marked extent where they extend forwards between the postero- 
lateral arms of the mesethmoid (Text-figs. 16, 17 and 31). Rhynchodercetis differs 
very slightly from the others in that the anterior rostral region is formed solely by 
the premaxillae and there is a tendency for the dentary to be shorter than the 
rostrum (Text-fig. 29), the latter being a feature seen in several other unrelated 
groups, e.g. Aspidorhynchus and Belonostomus (Gardiner, i960 : 362) among fossil 
halecostomes, and Istiophorus among Recent fish. There is obviously some correla- 
tion between the mode of life, particularly trophic habits, and the greater extension 
of the upper jaw, but any obvious correlation is difficult to envisage. Possibly the 
rostrum acted as a cutwater, reducing the amount of turbulence during forward 
motion and thus enabling the fish to approach its prey undetected. All the other 



ESPECIALLY MYCTOPHOIDS 177 

genera are also predatory as evidenced by the well-developed dentition on all the 
jaw bones. The jaw suspensorium remains vertical in position irrespective of the 
lengthening of the jaws and is always shallow. This factor alone tends to indicate 
that these genera moved on to prey slowly and undetected. A fast movement 
would be hazardous to the long thin jaws which would have had to have been opened 
during this rush. 

Bodily elongation occurs within the Dercetidae as a result basically of the multi- 
plication of vertebral elements (as many as 80 being present in Dercetis gracilis, 
Text-fig. 25). Additional lengthening is achieved by extension of the vertebral 
centra in the anterior precaudal region in both Dercetis and Rhynchodercetis. The 
outcome of this extension of the centra is a fish with a very long and narrow ' neck ' 
region, well shown in Dercetis rostralis (Text-fig. 26), in which the anterior centra are 
as much as six times as long as they are deep. Together with the extension of the 
' neck ' there is the production of enlarged transverse processes projecting strongly 
laterally (Text-fig. 30A). Siegfried (1966 : 214) has compared the dercetids with the 
Gasterosteiformes in respect of the vertebral composition. Elongation of the 
anterior precaudal vertebrae occurs in the Aulostomoidei (Aulostoma) and the 
Syngnathoidei (Syngnathus) , and in these genera large transverse processes occur. 
The gasterosteiforms are in no way related to the dercetids, since the elongate 
dercetid jaw is at complete variance with the very small jaws at the end of a long 
tubular snout encountered in the Gasterosteiformes. 

In Rhynchodercetis hakelensis (Text-fig. 28) long, linearly arranged ossifications in 
the mid-dorsal line between the occiput and the origin of the dorsal fin may represent 
ossifications within a median dorsal ligament. If the transverse processes were in 
life connected one to another by ligaments, then together with the ossified dorsal 
ligament, the ' neck ' could have been held rigidly in a straight line. Since the 
median fins occur in the posterior half of the body and the caudal region is relatively 
stout, then this may support the assumption that these fish moved slowly and 
stealthily towards their prey, the anterior body being held rigid and thus creating 
little turbulence in the water, whilst the main propulsive force is concentrated at 
the extreme posterior end of the body. On the other hand the modifications of the 
body may have enabled the fish to make a quick darting movement, this however 
has been shown not to be in accord with the jaw structure. Arambourg (1954 : 119) 
stated that Rhynchodercetis yovanovitchi resembled the Scomberesocidae and certain 
Xiphiidae, and was probably exclusively pelagic. This would seem to be a reason- 
able conclusion and the habit of the gar-fish, Belone, is closest to that of the dercetids. 

Squamation in the Dercetidae is reduced to several isolated scute rows along the 
flanks, the major rows being composed of scales conforming to a basic triradiate plan 
(Text-figs. 82A-D). In general, reduction in squamation is correlated with the 
presence of a distensible stomach, and this is so in the dercetids, since several 
specimens examined contained whole fish within the body cavity (see also Woodward, 
1942a : 555). 

The genus Prionolepis has so far not been mentioned in the systematic descriptions 
or in the discussion. This genus, originally refered to the Enchodontidae by 



i 7 8 



UPPER CRETACEOUS TELEOSTS 



Woodward (1901 : 230) presents some problems as to its relationships. It would 
seem to be most closely related to the dercetids and cimolichthyids as constituted in 
this work. Provisionally then it is placed as an addendum to the suborder Cimo- 
lichthyoidei and a new family Prionolepididae is created for its reception. There 
follow diagnoses and a description of the only completely preserved species, 
Prionolepis cataphractus. 



A 





\ \ 






Fig. 82. Flank scutes of representative cimolichthyoids. (a) Dercelis triqueter Pictet. 
(b) Dercetis gracilis (Davis), (c) Rhynchodercetis gortanii (d'Erasmo). (d) Pelargor- 
hynchus dercetiformis von der Marck. 



Family PRIONOLEPIDIDAE nov. 

Diagnosis. Body elongated and relatively shallow, neurocranium shallow and 
acutely pointed anteriorly. Post-temporal fossa roofed. Mesethmoid with large 
ventro-laterally directed, wing-like structures. Supratemporal sensory canal on the 
parietals and supratemporals. Maxilla stout with two prominent rows of teeth ; 
supramaxilla absent. Vertebrae approximately 45 in number of which at least 20 
are caudal. Dorsal fin short and in the middle of the back ; anal fin small and 
remote. Body bearing a single row of deep scutes which cover the whole flank 
region on either side. 



ESPECIALLY MYCTOPHOIDS 179 

Genus PRIONOLEPIS Egerton, 1850 

Diagnosis (emended) . Prionolepididae in which the head and body are slightly 
elongated and somewhat laterally compressed. Teeth extremely numerous on the 
vomer, palatine, ectopterygoid and dentary ; all teeth long, thin and needle-like 
and noticeably recurved. Vertebrae as long as deep with mesially constricted 
centra. Paired fins large, pectorals slightly larger than the pelvics. Pelvic fins 
opposed to the hinder part of the dorsal fin. 

Type species. Prionolepis angustus Egerton. 

Prionolepis cataphractus (Pictet & Humbert) 
(Text-figs. 83-85) 

1866 Aspidopleurus cataphractus Pictet & Humbert : 109, pi. 18, fig. 1. 
1901 Prionolepis cataphractus (Pictet & Humbert) Woodward : 230. 
Woodward lists the earlier references. 

Diagnosis (emended). Prionolepis of standard length not exceeding 19 cm. 
Length of head with opercular apparatus less than half the length of the body. 
Maximum depth of the trunk one-seventh of the body length. Dorsal fin with 18 
rays ; anal fin with 14 rays, nearer to caudal than to pelvics. Pelvic with 9 rays, 
pectoral with 13 rays. Lateral scutes 45 in number, the largest being four times as 
deep as broad. Hinder margin of each scute slightly pectinate, anterior margin 
marked with delicate and irregularly arranged vertical striations. 

Holotype. Specimen number 639/59, an imperfect fish in the Museum d'Histoire 
Naturelle, Geneva, from the Middle Cenomanian of Hakel, Lebanon. 

Material. Specimens in the B.M.N.H., numbers P. 4864, P. 9966, P. 9967, 
P. 9968, P. 9970, P. 47332, P. 47516, all from the Middle Cenomanian of Hakel and 
Hajula, Lebanon, and all prepared by the transfer method. 

Description. Neurocranium. The neurocranium is shown in dorsal view in 
Text-figure 83. The skull-roof is flattened and reaches its widest extent between the 
sphenotics at the hind end of the orbit. The frontals occupy the majority of the 
roof extending back almost to the occiput. The frontals contact each other in a 
straight suture. Laterally the frontals form the upper border of the orbits and 
extend on to the dorsal surface of the sphenotics. More posteriorly the frontals 
contact the medial edges of the pterotics in a straight suture. The frontal is 
ornamented with raised ridges of bone each set with minute tubercles. The orna- 
mentation of the frontal arises from the centre of ossification above the hind end of 
the orbit. The ridges pass forwards towards the snout region, and also laterally 
above the orbit. Posteriorly directed ridges pass back on to the rear part of the 
frontal above the cranial cavity. Anteriorly the frontals taper most markedly from 
the anterior orbital region and are inserted below the posterior edge of the meseth- 
moid. Posteriorly the frontals overlap the anterior regions of both the parietals 
and the supraoccipital. 



i8o 



UPPER CRETACEOUS TELEOSTS 



The supraoccipital contacts the frontals anteriorly, and separates the parietals. 
It extends down on to the posterior face of the neurocranium where it gives rise to a 
small backwardly directed crest which does not extend above the plane of the skull- 
roof. Laterally the supraoccipital meets the epiotic in a pronounced suture. 




nf.s.c 



inf. s.c 



popsc 



st. s.c 



Fig. 83. Prionolepis catapkractus (Pictet & Humbert). Neurocranium in dorsal view. 
The broken line on the right side of the figure indicates the course of the sensory canals. 



ESPECIALLY MYCTOPHOIDS 181 

The parietals are bordered anteriorly by the frontals and posteriorly by the 
epiotics. The parietal is smooth apart from two small strips of ornamented bone 
fused to the surface. One of these strips is transversely arranged, the second 
longitudinally. Both the transverse and longitudinal strips contain partially roofed 
dorsal grooves. These two strips of bone represent the remains of a supratemporal 
bone, and housed the supratemporal sensory canal. 

The pterotic forms the postero-lateral border of the skull-roof contacting the 
sphenotic anteriorly and the frontal medially. Postero-medially it is separated 
from the parietal and the epiotic by an indentation of the posterior edge of the skull- 
roof. This indentation is in the form of a narrow antero-posteriorly orientated 
groove, probably representing the upper region of the opening to a roofed post- 
temporal fossa. The dorsal surface of the pterotic is ornamented with irregular, 
tuberculated ridges. A tube is present within the pterotic opening anteriorly about 
mid-way along the bone, and posteriorly in two places. The first of these posterior 
openings is positioned on the edge of the pterotic and the second on the inner edge 
of the posterior part of the pterotic. On the lateral face of the neurocranium the 
pterotic makes up the major part of the hyomandibular facet. Above the hyo- 
mandibular facet the pterotic contributes towards the formation of the dilatator 
fossa which is an elongated oval fenestra. 

The sphenotic projects laterally from beneath the frontal at the rear of the orbit. 
Dorsally the sphenotic joins the pterotic posteriorly, and is unornamented. 
Laterally the sphenotic gives rise to a stout strut passing ventro-medially to contact 
a corresponding strut of the prootic. Between these two struts a cup-shaped recess 
is produced and this forms the anterior region of the hyomandibular facet. 

The course of the sensory canals is indicated both by tubes within the bone and 
open grooves on the surface. The supraorbital sensory canal was confined to the 
frontal, and opened anteriorly on the edge of the frontal in advance of the orbit : 
no trace of nasal bones was found in this region. Posteriorly the sensory canal 
passed postero-medially from the centre of ossification. The infraorbital sensory 
canal passed across the dorsal surface of the sphenotic and under the lateral extension 
of the frontal. The lateral line canal from the post-temporal appears to have 
divided before entering the posterior part of the neurocranium. The otic branch 
of the infraorbital canal entered the pterotic above the post-temporal fossa and 
passed forwards within the bone. Anteriorly the otic branch passed on to the 
dorsal surface of the sphenotic beneath the frontal to contact the main infraorbital 
canal. The supratemporal sensory canal is the second division of the main lateral 
line canal, and passed medially on to the surfaces of both of the limbs of the supra- 
temporal. The supratemporal canal appears to have been continuous across the 
dorsal surface of both the parietals and the supraoccipital. The preopercular 
sensory canal was transmitted through a postero-lateral pore of the pterotic and 
joined the otic branch of the infraorbital canal within the pterotic. 

The epiotics form part of the occipital border and the medial edge of the posterior 
part of the post-temporal fossa. The major portion of the epiotic is on the posterior 
face of the neurocranium. Postero-dorsally the epiotic gives rise to a smooth, 



182 



UPPER CRETACEOUS TELEOSTS 



rounded, posteriorly directed flange. The dorsal surface of this flange provides the 
region of articulation for the dorsal limb of the post-temporal. 

The mesethmoid is peculiarly shaped, appearing as two distinct portions, an 
anterior short, straight rod-like piece forming the extreme anterior termination of 
the snout ; and a much expanded posterior region. This posterior region is in the 
form of two narrow, flattened wing-like expansions which are directed postero- 
laterally and inclined ventro-laterally. The lateral edge of each mesethmoid wing 
is attached to the dorso-medial region of each palatine. 

The vomer closely parallels these lateral extensions of the mesethmoid and is 
attached to its ventral surface. The vomer is more extensive than the mesethmoid 
and gradually narrows posteriorly to contact the parasphenoid in the region of the 
lateral ethmoid. The palatine appears to be associated with the edge of the vomer 
which supports numerous teeth. 

The parasphenoid is visible within the orbit where it is slightly expanded into 
lateral flanges which overlap the medial edges of the endopterygoids. Anteriorly the 
parasphenoid passes between the lateral ethmoids and attaches to the posterior end 
of the vomer. The lateral ethmoids meet the ventral region of the frontals delimiting 
the orbit anteriorly. They expand ventrally and are attached to the upper posterior 
end of the palatine. 

Hyopalatine bones. The hyomandibular is large with an elongated head. The 
anterior part of this head is thickened and so shaped as to fit into the anterior cup- 
shaped part of the hyomandibular facet. The remainder of the head is much thinner 
and almost straight. The posterior edge of the hyomandibular is backwardly 
curved, but the posterior edge of the quadrate curves forwards so that the condyle 
occurs immediately below the occiput. The outer face of the hyomandibular bears 
one relatively prominent crest arising from the anterior region of the head and 
passing postero-ventrally to the hind edge of the bone. 




Fig. 84. Prionolepis cataphractus (Pictet & Humbert). Restoration of the skull in 

lateral view. 



ESPECIALLY MYCTOPHOIDS 183 

The quadrate is a large, triangular bone with a prominent transversely arranged 
condyle. The posterior edge of the quadrate is thickened and deflected laterally 
and bears a vertical groove posteriorly. The postero-dorsal edge of the quadrate is 
slightly excavated, indicating the upper limit of the symplectic groove on its medial 
face. The symplectic is a vertical rod-like bone, internal to the quadrate. The 
anterior end of the quadrate lies against the lateral surface of the posterior end of the 
ectopterygoid. 

The ectopterygoid is a long narrow strut of bone slightly thickened posteriorly. 
The ventral edge of the ectopterygoid is flattened and bears teeth. Anteriorly the 
bone has some three rows of slender, acutely pointed recurved teeth, but more 
posteriorly the tooth row number decreases. The ectopterygoid contacts the 
palatine anteriorly and is quite firmly attached to it. The endopterygoid is plate- 
like and attached to the internal edge of the ectopterygoid. The metapterygoid is a 
very large thin sheet of bone covering most of the lateral face of the hyomandibular, 
contacting the quadrate and ectopterygoid ventrally and the endopterygoid 
anteriorly. 

The palatine is elongate, stout and forms a direct continuation of the ectopterygoid 
anteriorly. Postero-dorsally it is attached to the ventral region of the lateral 
ethmoid, and anterior to this has a small dorso-laterally inclined maxillary process. 
The palatines are very close together at the anterior end of the snout. Ventrally the 
palatine is flattened and covered irregularly by teeth. The teeth are slender, 
acutely pointed and strongly recurved. 

Dermal upper jaw. The dermal upper jaw is shown in lateral view in Text-figure 
84. The premaxilla is prominent, thin and elongate. Anteriorly it appears to 
contact its partner in the mid-line just in advance of the mesethmoid. The outer 
face of the premaxilla is smooth and curved. It is of relatively uniform depth along 
most of its length but tapers somewhat posteriorly. 

The maxilla is long and slender, forming almost exactly half of the gape. The 
head of the maxilla is inflected and bears a medial excavation which houses the 
maxillary process of the palatine. The anterior region of the maxillary head is 
drawn out in front of this medial excavation and would appear to lie on the dorsal 
surface of the wing-like mesethmoid extension. 

Teeth are present on both of the bones in the dermal upper jaw. The maxilla 
bears a double row of small acutely pointed teeth on that part of the bone which 
enters the oral border of the upper jaw behind the premaxilla. The teeth are 
inclined slightly in a posterior direction. Teeth on the premaxilla are confined to 
a single marginal row and are essentially similar to those on the maxilla but smaller. 

Mandible. The mandible is shown in lateral view in Text-figure 84. It is long, 
very shallow and acutely pointed at the symphysis. The dentary forms the whole of 
the oral border and practically the entire ventral border and has a deeply indented 
posterior edge. The oral margin of the dentary supports numerous irregularly 
placed, small teeth. The teeth are acutely pointed and markedly recurved. 
Towards the posterior end of the dentary the teeth are reduced in number and 
confined to approximately two rows. 



184 UPPER CRETACEOUS TELEOSTS 

The articular forms the posterior region of the mandible, extending anteriorly to 
insert on the medial face of the dentary. Below the shallow articular facet there is a 
small posteriorly directed foramen which transmitted the mandibular sensory canal. 
This sensory canal passed within a tube of bone on the ventro-lateral edges of both 
the articular and dentary, and a small oval pore close to the symphysis indicates its 
anterior extent. The lateral face of both dentary and articular is unornamented, 
but the articular has a pronounced lateral flange passing forwards from in front of 
the articular facet. This flange decreases in size and ends near to the level of the 
posterior end of the premaxilla. The maxilla rests against the dorsal surface of this 
flange when the jaws are shut. 

The angular is a small knob of bone forming the postero-ventral angle of the 
mandible, with its posterior face fitting against the ventral region of the pre- 
operculum. 

Opercular bones. The opercular bones are shown in lateral view in Text-figure 84. 
The preoperculum is deep, and narrow. Dorsally it ends posterior to the head of the 
hyomandibular and anteriorly lies against the posterior face of the crest on the hind 
edge of the hyomandibular. Ventrally the preoperculum rests in a groove on the 
posterior edge of the quadrate, where it is slightly expanded and bears several promi- 
nent ridges. The preopercular sensory canal ran within a tube along the length of 
the bone opening ventrally through three or four pores. The most prominent pore 
faces antero-ventrally and transmitted the main canal to the mandible. The 
posterior edge of the preoperculum is thin and overlaps the anterior edge of the 
operculum. 

The operculum is large and the opercular facet occurs just within the upper half of 
the anterior edge. The posterior edge of the operculum is rounded dorsally, but 
ventrally the edge is irregularly excavated. A horizontal strengthening ridge passes 
across the internal face of the operculum. 

The suboperculum lies ventral to the operculum and is overlain by the latter to a 
marked extent. The supoberculum has a smooth outline, unlike the operculum, 
and its lateral face is unornamented. No trace of an interoperculum was found in 
any of the specimens examined. 

Hyoid arch. Both the ceratohyal and the epihyal are elongate, roughly rectangular 
ossifications, joined in life by cartilage. A prominent groove runs along the length 
of the lateral face of both bones and marks the course of the hyoidean artery. The 
hypohyals are prominent and attached to each other in the mid-ventral line. The 
ceratohyal is firmly attached to the posterior end of the hypohyal, and traces of a 
median urohyal are seen where it contacted the hypohyals. The urohyal is elongate 
with a shallow dorsal longitudinal crest. 

Fourteen branchiostegal rays are present, the first 8 attached to the ventral region 
of the lateral face of the ceratohyal. These 8 rays are uniform in size and shape. 
The more posterior 6 branchiostegal rays articulate with the epihyal and are stouter, 
slightly flattened and more sharply inclined posteriorly. 



ESPECIALLY MYCTOPHOIDS 185 

Pectoral girdle and fin. The post-temporal has a flattened, oval dorsal limb 
articulating with the epiotic. The upper surface of the dorsal limb is ornamented 
with bony tubercles postero-medially. The ventral limb is in the form of a narrow 
antero-ventrally inclined strut. The supracleithrum articulates with a postero- 
medial facet on the post-temporal. The lateral line canal entered the post-temporal 
through a foramen in the supracleithral facet and passed through the post-temporal 
to emerge near its anterior end. 

The supracleithrum is short and stout with an expanded, elongated, head. Dorso- 
laterally it is ornamented with minute bony tubercles and more ventrally with 
raised ridges. The lateral line canal entered the posterior region of the head of the 
supracleithrum and passed anteriorly within the bone. 

The cleithrum is a large curved bone narrowing dorsally where it fits under the 
supracleithrum, and curving forwards ventrally. The anterior edge of the cleithrum 
is inclined medially forming the posterior boundary of the branchial opening. 

The endoskeletal girdle projects behind the cleithrum, and a large scapular 
foramen is visible. On the posterior edge of the scapula a large saddle-shaped facet 
forms the articulation for the anterior fin ray. The coracoid is thinner and more 
membranous than the scapula and extends forwards ventrally to contact the 
cleithrum at its tip, thus leaving a large fenestra between the two. 

Two postcleithra are present, the dorsal one being oval in shape, and attaching to 
the medial face of the cleithrum and supracleithrum. The ventral element is thin 
and curved and extends ventrally, medial to the pectoral fin. 

The pectoral fin is composed of 13 rays all of which are segmented and branched 
distally, except for the first which is unbranched. 

Pelvic girdle and fin. Each pelvic bone is an elongate triangle lying horizontally 
in the ventral body-wall about halfway along the length of the body, the insertion 
of the fin being on a level with the 21st precaudal vertebra. There is a thickened 
ventral flange running the length of each pelvic bone and anteriorly the bones are 
joined. Postero-laterally the pelvic bone is thickened to form a lateral condyle. 
Medial to the condyle the remainder of the fin rays articulated with an elongated 
facet. Medial again to this facet the pelvic bone contacts its partner in an irregular 
interdigitating suture. Each fin consists of 9 rays, all except the first being branched, 
and all segmented. 

Vertebral column. The vertebral column is shown in the restoration, Text-figure 
85. It consists of 45 vertebrae of which approximately 21 are caudal. 
The centra are almost twice as long as deep and slightly mesially con- 
stricted. Dorsally each centrum bears a neural arch and spine, and those of the 
anterior precaudal vertebrae are laterally compressed and slightly expanded. The 
neural arches represent separate ossifications from the centra ; the suture between 
arch and centrum is clearly visible along the entire length of the column. The 
precaudal vertebrae bear small transverse processes ventro-laterally on the anterior 
region of each centrum, with which pleural ribs articulate. The ribs are long, 
slender, tapering and inclined posteriorly. Each haemal arch is drawn out into a 



1 86 



UPPER CRETACEOUS TELEOSTS 




a 




o 




4-> 




_0 








o 




M 




en 




<x> 




sz 


dJ 






4H 


"re 


o 


0) 


c 


V 


_0 


c 


+J 


;*~] 


oi 




Ih 
O 

OI 


u 
ctj 




cu 




x; 




+j 


1) 


O 


,D 


,_l 


£ 


ctj 


3 


cu 


I 


> 




■r. 


°8 


en 


■p 


? 


dj 


O 


o 





to 





CD 

x; 


»-52 


H 


§, 




« 








"5 


-d 


v> 


a- 




+j 


<o 


+j 








£ 


o 


o 


S 


en 


o 


dj 






£ 


cfi 


^ 


o 
en 


>o 




00 




6 








to 





ESPECIALLY MYCTOPHOIDS 187 

backwardly projecting haemal spine. Intermuscular bones, both epineurals and 
epipleurals, are present on the first 26 vertebrae. 

Median fins and tail. The median fins are shown in the restoration, Text-figure 85. 
The dorsal fin consists of approximately 18 rays and occurs about mid- way along the 
length of the body. The first ray is unbranched and shorter than the subsequent 
rays which are all branched and segmented. The proximal radials are elongate 
and the anteriormost are expanded, especially the first which bears a large forwardly 
directed keel-like expansion. The heads of all the proximal radials are expanded. 
Certain of the narrow medial radials are visible extending horizontally between 
the bases of adjacent fin rays. 

The anal fin is made up of 14 rays and positioned at the extreme posterior end of 
the body close to the caudal fin. As in the dorsal fin the first ray is unbranched. 
The proximal radials are prominent and elongate but not expanded. Medial radials 
occur between the bases of adjacent fin rays. 

The caudal skeleton is composed of 6 vertebrae. There are three free preural 
vertebrae (preurals 2, 3 and 4) ; one fused preural and ural vertebra (preural 1 and 
ural 1) ; and one free ural vertebra (ural 2) which is represented by a terminal half- 
centrum. Preurals 3 and 4 bear normal neural and haemal spines, whereas preural 
2 bears an expanded haemal spine and only a neural arch, the neural spine having 
been lost. The compound vertebra (preural 1 -f ural 1) bears the parhypural 
(haemal spine of preural 1) and two hypurals, the first of which is the larger. Ural 
vertebra 2 has approximately three hypurals associated with it making at least five 
hypurals in all. Associated with the dorsal regions of the ural vertebrae are two 
uroneural elements, which are elongated in a postero-dorsal direction. At least two 
epural elements are present, one of which is expanded. 

Nineteen principal fin rays are present of which 17 are branched. Accessory fin 
rays are present above and below the fin and number at least 20. 

Scales. Several of the scales are shown in the inset to Text-figure 85. They are 
present as a single continuous row along the lateral region of the body, and represent 
much enlarged lateral line scutes. The scales begin immediately behind the supra- 
cleithrum and terminate on the caudal peduncle. The row consists of 45 scutes, the 
most anterior and posterior ones being the smallest. Most of the scutes are about 
three times as deep as they are broad and the lateral line canal ran within their 
centres. The scales overlap and posteriorly are irregularly serrated. In the region 
of the lateral line there is a somewhat deeper indentation of the rear margin which is 
overhung by a backwardly projecting spine. The anterior edge of each scale is 
thickened and dorsally and ventrally the edges are rounded. The scales are 
obliquely inclined and ornamented with ridges on their exposed portions. The 
posterior part of the lateral face bears horizontally arranged ridges, whilst the 
anterior regions have vertically arranged ridges. Towards the caudal peduncle the 
scales become reduced in depth but the backwardly projecting spines become more 
prominent so that pronounced lateral flanges are produced on either side of the tail. 

From the above description it can be seen that Prionolepis does share certain 
common characteristics with the Cimolichthyoidei. The skull roof is similar, with 



188 UPPER CRETACEOUS TELEOSTS 

the roofed post-temporal fossa ; the small, discrete laterally facing dilatator fossa 
above the hyomandibular facet ; toothed vomer ; simple nature of the premaxilla ; 
long narrow maxilla which rests on a shelf on the lateral face of the mandible when 
the mouth is closed ; vertical jaw suspension ; absence of an interoperculum ; 
presence of numerous intermuscular bones, and abdominal pelvic fins. 

The rostral region of Prionolepis is somewhat lengthened, much as in Cimolichthys, 
but the body is little elongated. One of the most distinctive features of Prionolepis 
is the mesethmoid arrangement (Text-fig. 83). The anterior rod-like portion with 
which the premaxillae associate and the two large postero-laterally arranged 
flattened wings are different to the condition of this region in either Cimolichthys 
levesiensis or the dercetids. However in Cimolichthys nepaeolica from the Niobrara 
Chalk of Kansas, the mesethmoid arrangement approaches the condition in Priono- 
lepis. As with Cimolichthys levesiensis (p. 40, Text-fig. 16), Cimolichthys nepaeolica 
(Hay, 1903, fig. 69) has the same large backwardly divergent laminae, but anteriorly 
there is a forwardly projecting prominence closely resembling the straight rod-like 
piece in Prionolepis cataphr actus (p. 182, Text-fig. 83). This mesethmoid arrange- 
ment appears to be distinct from all other teleost groups and provides good evidence 
for the relationship between Cimolichthys and Prionolepis. 

There are no expanded transverse processes on the precaudal centra, in this 
respect differing markedly from the Dercetidae. Cimolichthys, however, also has 
no such expanded processes although it is obviously closely related to the dercetids. 
The deepened, shield-like flank scutes in Prionolepis are of interest (Text-fig. 85) 
since they show signs of being feebly ctenoid. However the scales look more like 
the pectinate scales of Lepidotes rather than true ctenoid scales. The deepened 
flank scutes recall those of Aspidorhynchus and Belonostomus (Gardiner, i960), and 
among recent fish similar scutes are exhibited in the Gasterosteiformes. It is of 
interest to note that in the gasterosteoids the flank scutes are supported on large 
transverse processes. 

The systematic position of the Cimolichthyoidei is not at all clear. Pictet (1850) 
originally used the family Dercetidae (Hoplopleuridae) to include the genera 
Dercetis, Leptotrachelus, Pelargorhynchus, Saurorhamphus and Eurypholis, being 
those fishes devoid of normal squamation but possessing isolated scute rows. 
Woodward (1901 : 171) used Pictet's family Dercetidae but removed Saurorhamphus 
and Eurypholis to the Enchodontidae. Since 1901 one further genus has been 
added by Arambourg (1944 : 281), Rhynchodercetis. 

The dercetids in the past have been associated with one or another of two major 
groups. Woodward (1901) related the dercetids to the halosaurs and notacanths. 
Jordan (1923 : 128) more or less concurred and grouped the Dercetidae (now includ- 
ing Prionolepis and Leptecodon) in the order Heteromi stating that they were close to 
the notacanthids. 

The more predominant line of thought, however, has been that the dercetids are 
more nearly related to the myctophiform assemblage. Regan (191 1 : 120) stated 
that the orbital, postorbital and post-temporal parts of the Dercetidae resembled 
Evermannella, whilst the ethmoid region and jaws resembled Alepisaurus. 



ESPECIALLY M YCTOPHOIDS 



189 



Gregory (1933 : 201) placed the dercetids in the Heteromi but stated that any 
similarities with the notacanths must be put down to convergence. Authors since 
that time (Berg, 1940 ; Bertin & Arambourg, 1958 ; Romer, 1966) have retained 
the Dercetidae within the Myctophiformes following Regan's (191 1) tentative 
proposals. 

The association of the dercetids with the notacanths is obviously incorrect due to 
several basic dissimilarities, for example the complete regression of the caudal fin 
skeleton, the small inferior mouth and the snubbed rostrum of the notacanths are 
vastly different to those of the dercetids. Greenwood et at. (1966) have placed the 
notacanths in the Elopomorpha, and following Arambourg (1954) placed the 
Dercetidae in the myctophiforms. 

The Cimolichthyoidei would appear not to be related to the myctophiforms 
because of certain primitive features which are not seen in the basal myctophiform 
stock, for instance, the absence of an ascending process and the toothed maxilla 
with no supramaxillae. 

The Cimolichthyoidei as a group were relatively numerous, especially the smaller 
dercetids, and extended throughout the Upper Cretaceous (except for the Maastrich- 
tian). The first genus to appear in the fossil record is the advanced Rhyncho- 
dercetis, in the Lower Cenomanian. The less advanced genera Cimolichthys and 
Dercetis do not appear in the record until the Upper Cenomanian. Thus the small 
cimolichthyoid radiation must have originated within the Lower Cretaceous. The 
advanced characteristics shown by the group exclude them from having given rise 
to any other salmoniform. These advanced characteristics are as follows : 

1. Parietals separated in the mid-line. 

2. Orbitosphenoid, basisphenoid and antorbital absent. 

3. Interoperculum absent. 

4. Reduced squamation. 

5. Abnormal transverse processes and lengthened centra. 

6. Overall snout and body elongation. 

Several of these advanced features also preclude the cimolichthyoids from having 
given rise to the Myctophiformes. The Cimolichthyoidei also retain many primitive 
characters in common with other members of the Salmoniformes such as : 

1. Partially or completely roofed post-temporal fossa. 

2. Maxilla enters the gape and is toothed. 

3. No ascending process on the premaxilla. 

4. Supratemporal present. 

5. Teeth on vomer and endopterygoid. 

6. Supraorbital present. 
Complex trigemino-facialis chamber. 
Abdominal pelvic fins. 
Epineurals in abundance. 

Despite the fact that the cimolichthyoids are in no way related to the myctophi- 
forms the two groups show a considerable degree of parallel evolution. For example 



igo UPPER CRETACEOUS TELEOSTS 

the caudal skeleton in the Cimolichthyoidei, as represented by Prionolepis, is 
practically identical to that of the Cretaceous myctophiform Sardinioides (Text-fig. 
72) and the Recent Aulopus (Text-fig. 93). In both the Cimolichthyoidei and the 
basal myctophiforms the first preural vertebra is fused with the first ural vertebra ; 
ural vertebra two is present as a small half-centrum supporting hypurals 3, 4, 5 and 6 ; 
a stegural is formed from uroneural one and the neural arches of ural vertebra one and 
preural vertebra one ; the neural spine of preural vertebra two is reduced so that only 
an expanded neural arch occurs. However a primitive feature of the myctophiform 
tail is the presence of basal fulcral scales both above and below the peduncle, these 
are absent in the cimolichthyoids. Apart from the tail other comparative points 
can be made between the cimolichthyoids and the more advanced ' alepisauroid ' 
myctophiforms. For instance the parietals are separated by the small supra- 
occipital in both groups. Nasal bones are absent in the cimolichthyoids, either 
having been lost, or more likely incorporated into the mesethmoid as a result of the 
lengthening of the rostrum. This fusion of the nasal with the mesethmoid also 
occurs in several advanced myctophiforms, e.g. Lestidium, Anotopterus, Alepisaurus 
(Parr, 1929 : 28, 34). Further the cimolichthyoid jaw resembles that seen in the 
' alepisauroid ' myctophiforms, except for the exclusion of the maxilla from the gape 
in the latter. In the cimolichthyoids (Text-fig. 21) the premaxilla does extend back 
for a considerable length below the maxilla, but never has an ascending or an articular 
process. The maxilla is reduced to a narrow strut of bone entering the gape poster- 
iorly, and never has articulatory head processes. In Alepisaurus (Text-fig. 79) on 
the other hand, the upper jaw, although at first sight simple and similar to that of 
the cimolichthyoids, has been derived from a much more complex condition. In the 
earlier myctophiforms the premaxilla possessed an articular and an ascending 
process, while the maxilla had a differentiated head and supported two supra- 
maxillae. As a result of specialization the articular process and the supramaxillae 
have been lost and the maxillary head is simplified. As in the cimolichthyoids this 
specialization is concerned with the lengthening of the rostral region. Other than 
this the cimolichthyoids still bear teeth on the vomer and the endopterygoids 
whereas in the advanced ' alepisaurs ' these teeth have been lost. 

In the cimolichthyoids barbed teeth occur on the palate and the ectopterygoid 
(Text-fig. 20) and serve to increase the total cutting edge of the teeth. Similar 
teeth are seen on the glossohyal (tongue) of Scopelarchus, but not elsewhere in any 
other myctophiform (Parr, 1929). 

No traces of an interoperculum have been found in the cimolichthyoids. This 
feature is difficult to account for but might in some measure be due to the retention 
of an upright suspensorium, since the interoperculum is often small in forms with a 
vertical preoperculum. In Alepisaurus (Text-fig. 79) and Omosudis the suspen- 
sorium is upright and the interoperculum minute, possibly indicating a stage in the 
disappearance of this bone. 

The transverse processes in the dercetids project strongly in a lateral direction, 
while Synodus among the recent myctophiforms has transverse processes which 
project ventrally. Two pairs of transverse processes are present on each centrum in 



ESPECIALLY MYCTOPHOIDS 191 

Synodus, as in Rhynchodercetis (Text-fig. 30A). What particular function they serve 
in Synodus is difficult to imagine. Rosen (1964 : 255) has pointed out the presence 
of many long, rod-like ossifications in the mid-dorsal line of Synodus anterior to the 
dorsal fin. Thus an analogous situation is found to that in Rhynchodercetis, in that 
these ossifications in Synodus may help to keep the body rigid. Synodtis is a ben- 
thonic form which rests on the bottom waiting for its prey. Probably the anterior 
region of the body is held rigid while it makes a quick darting movement towards its 
prey. Gosline, Marshall and Mead (1966 : 7) indicate that Synodus is capable of 
burying itself in sand, and the rigidity may be of use here. 

Reduction of the squamation is common among the most advanced myctophi- 
forms which possess a distensible stomach. There is no equivalent reduction to the 
several isolated scale rows seen in the Cimolichthyoidei ; lateral line scales may 
remain, however, on otherwise naked forms. 

The few points discussed above have illustrated that the cimolichthyoids are 
neither halosaurs nor myctophiforms, although similarities are shown. The Cimo- 
lichthyoidei most conveniently fit into the order Salmoniformes (sensu Greenwood, 
et. al., 1966), and represent a minor radiation which has evolved along similar lines 
to those within the Myctophiformes. Like the Myctophiformes the Cimolichthyoidei 
belong within the superorder Protacanthopterygii (sensu Greenwood, et. al., 1966). 

In conclusion the possession of a long snout and a long body would not seem to be 
the ideal prerequisites for a highly successful predator. Extant predatory forms 
possessing these characters are few in number, in fact the Belonidae would seem to 
be the only group (Bertin & Arambourg, 1958). Thus, in Recent fish, there seems 
to be little scope for a specialization of this kind to occur, since it would not result in 
as efficient or versatile a predatory form as a less specialized acanthopterygian. 
This factor might in part account for the disappearance of the Cimolichthyoidei once 
the radiation of the beryciforms and perciforms got under way at the end of the 
Cretaceous. 

(C) Suborder ENCHODONTOIDEI 

The first attempt to collect the enchodonts into a distinct group was made by 
Woodward (1901 : 189) when he erected the family Enchodontidae within the 
Isospondyli. This family he divided into two main sections dependent on whether 
a single terminal palatine tooth was present or not. In the first section (those with 
an enlarged terminal tooth) he included Enchodus, Palaeolycus, Eurypholis and 
Saurorhamphus. In the second section he included Halec, Cimolichthys, Prionolepis, 
Leptecodon and Pantopholis. Since 1901 three further genera have been included 
within the family, these are Apateodus (Woodward, 1903), Volcichthys (d'Erasmo, 
1946) and Rharbichthys (Arambourg, 1954). 

The suborder Enchodontoidei as presented in this work only includes those genera 
which Woodward (1901) separated as possessing an enlarged palatine tooth. These 
four genera have been distributed between two families, the Enchodontidae con- 
taining Enchodus and Palaeolycus, and the new family Eurypholidae containing 
Eurypholis and Saurorhamphus. 



192 UPPER CRETACEOUS TELEOSTS 

The first obvious difference between the two families is the structure of the neuro- 
cranium and in particular the skull-roof (compare Text-figs. 32 and 44). In the 
Enchodontidae the roof inclines ventrally in its postero-lateral region due to the 
disappearance of the roofs of both the post-temporal and dilatator fossae. In most 
forms in which the roof of the post-temporal fossa is lost the reduction is correlated 
with the tendency for the body to become deepened post-cranially. The epaxial 
musculature of the ' neck ' region in deeper bodied fish passes antero-ventrally to 
insert on to the rear end of the skull, the skull-roof and into an unroofed post- 
temporal fossa. In shallow bodied fish the epaxial musculature passes directly 
anteriorly and attaches solely to the rear face of the neurocranium or into a roofed 
post-temporal fossa. This latter condition is seen in the second family, the 
Eurypholidae (Text-fig. 47), where the body is undeepened, the skull-roof flattened 
and the post-temporal fossa roofed. In the Enchodontidae the loss of the roof of the 
post-temporal fossa is associated with a general trend towards body deepening 
(Text-fig. 41). Probably correlated with this postcranial deepening is the tendency 
for the body to become laterally compressed. Throughout the Enchodontidae the 
body is stouter in relation to the head than it is in the Eurypholidae. Eurypholis 
boissieri (Text-fig. 47) has an enormous head compared to the body, and far from 
exhibiting any postcranial deepening the body appears slightly dorso-ventrally 
flattened. The overall neurocranial depth is greater in Enchodus than in Eurypholis, 
and the hyomandibular is deeper. Both of these factors are again correlated with 
the overall tendency for the deepening of the body in the Enchodontidae as opposed 
to the Eurypholidae. 

Body elongation has occurred in both of the families. In the Enchodontidae, 
Palaeolycus has a typical enchodontid head region (comparable to the genus 
Enchodus) but the body has become greatly extended by the multiplication of the 
vertebrae, so that upwards of 70 are present (Text-fig. 43). A corresponding 
elongation has occurred in the anal fin. Within the Eurypholidae, Saurorhamphus 
has an elongated body together with an elongated snout (Text-fig. 55). The body is, 
however, extended not by the multiplication of vertebrae but by the lengthening of 
the precaudal centra. 

The paired fins and associated fin girdles differ somewhat in the two families. 
The Enchodontidae possess a ' normal ' pectoral girdle with a relatively unexpanded, 
curved cleithrum and the pectoral fin is low on the flank (Text-fig. 40). The pelvics 
are abdominal, lying below the origin of the dorsal fin, with a tendency to be reduced 
in size somewhat. Palaeolycus has a minute pair of pelvics but the pectorals are 
enlarged. In contrast, the Eurypholidae have rather more specialized paired fins. 
The cleithrum is greatly expanded both anteriorly and posteriorly, and ventrally is 
turned in towards the mid-line (Text-fig. 46). The cleithrum is extensively 
ornamented and posteriorly extends below the fin insertion. The pectoral fin is 
much higher on the flank than it is in the Enchodontidae. The pelvic bones are 
large and stout and positioned directly behind the posterior cleithral extension. 
The pelvic fins themselves are thus in a sub-thoracic position and are larger than the 
pectorals. 



ESPECIALLY MYCTOPHOIDS 193 

Both families have a reduced squamation, scales being present only along the 
lateral line and in the mid-dorsal line anterior to the dorsal fin. In the Enchodon- 
tidae this squamation is further reduced and neither the mid-dorsal scutes nor the 
lateral line scales overlap. The mid-dorsal scutes are most prominent in the 
undeepened species, e.g. Palaeolycus dreginensis (Text-fig. 43), and are smallest in 
those species of Enchodus which have some post-cranial deepening, e.g. Enchodus 
marchesetii (Text-fig. 41). In the Eurypholidae both series of scutes are well 
developed, especially those in the mid-dorsal line. In Eurypholis boissieri there is 
evidence that a sensory canal passed through the mid-dorsal scutes (p. no). This 
median sensory canal presumably connected with the supratemporal commissure on 
the supraoccipital (Text-figs. 44 and 49). The lateral line scales are larger and more 
prominent in the Eurypholidae, but in both families the scales are of identical shape. 
Each scale bears a small pointed prominence. The prominence is smallest at the 
anterior end of the lateral line, but towards the caudal peduncle the spines become 
produced into larger hook-shaped projections. This increase in size is most marked 
in the Enchodontidae (Enchodus major) where the last lateral line scale bears a large, 
pointed horizontal flange projecting outwards and backwards. Monod (1959) 
noticed a similar latero-caudal spine in A canthurus monroviae formed from a modified 
scale. Monod found no intrinsic musculature in the spine and it did not secrete 
venom, but he could give no suggestions as to its function. It can only be suggested 
that the spine in Enchodus may have had some stabilizing effect during swimming, or 
possibly an offensive effect if the tail was thrashed through the water. 

These familial differences can possibly be related to a basic divergence in habitat 
and mode of life. Both families are clearly composed of predatory species with 
streamlined bodies, stout jaws, and a formidable armament of teeth. The Enchodon- 
tidae, with their slightly deepened, laterally compressed bodies, appear to have been 
active mid-water or pelagic predators. The Eurypholidae, on the other hand, with 
their lean, shallow, slightly dorso-ventrally compressed bodies, were probably 
bottom dwellers. The great expansion of the cleithra ventrally in the Eurypholidae 
possibly offered a protection against abrasion. In Eurypholis boissieri and 
Saurorhamphus freyeri the ventral edge of the mandible is strongly inclined medially 
and the mandibular sensory canal is enclosed within a tube for most of its length. 
In the Enchodontidae no such medial inclination of the mandible occurs and the 
sensory canal is housed in a shallow groove. Thus in the Eurypholidae both throat 
and sensory canal are protected from abrasion. The pelvic fins in the Eurypholidae 
are sub-thoracic and enlarged and possibly these may have served as supports on 
which the fish rested on the bottom. This last fact, that of waiting for the prey, is 
in part substantiated by the leanness of the body indicating some muscular reduction. 
Additional components of the sensory canal system on the dorsal surface of the 
Eurypholidae probably increased the perceptiveness to stimulation from above. 
Gosline, Marshall and Mead (1966 : 7) have considered a similar aspect in relation to 
the frontal commissure of the supraorbital canals in the Myctophiformes. In no 
other teleosts, living or fossil, can I find a mid-dorsal sensory canal like that of the 
Eurypholidae. 



194 UPPER CRETACEOUS TELEOSTS 

Despite these different modes of life the two families share a common jaw structure 
and mode of functioning. This aspect together with a discussion on the status of the 
ascending process of the premaxilla has been considered by the author in a separate 
paper (Goody, 1968). 

The relationships and affinities of the enchodontoids have not, hitherto, been more 
than scantily covered. Woodward (1901 : 189) placed the family within the 
Isospondyli close to the Scopelidae, indicating that their nearest living relatives were 
the Omosudidae and the Alepisauridae. In the introduction to volume 4 of his 
" Catalogue " (1901), Woodward stated that, " The Enchodontidae represent a very 
abundant family which might furnish the ancestors of both the Isospondylous 
Scopelidae or the Acanthopterygian Berycidae ". Regan (191 1 : 120) put the 
family in the Iniomi within the Isospondyli, corresponding roughly to Woodward's 
(1901) designation, but related them closer to the stomiatoids. Jordan (1923 : 126) 
followed Regan in relating them to the stomiatoids. Gregory (1933 : 204) placed 
the family within the group Mesichthyes (intermediate teleosts) , and pointed out the 
divergence of opinion between Woodward (1901) and Regan (191 1) as to their 
affinities, either scopeloid or stomiatoid respectively. Berg (1940 : 432) erected the 
new suborder Enchodontoidei within the Clupeiformes to contain the single family 
Enchodontidae, related to the stomiatoids. Bertin and Arambourg (1958 : 2268), 
realigned the family with the myctophiform fishes, putting them within the suborder 
Alepisauroidei. 

Previous opinion then, has been to associate the enchodontoids with either the 
stomiatoids or the myctophiforms, although general agreement has been reached in 
assuming the enchodontoids to be relatively primitive and close to the base of the 
teleost radiation. 

The stomiatoids are a pelagic or bathypelagic group of fishes of uncertain affinities. 
Beebe and Crane (1939) indicated that they were more closely related to the 
salmoniform fishes than to the clupeoids. This view has now been amply justified 
by Weitzman (1967), who has derived the stomiatoids from a stock close to the 
osmerids. Several features shown by the stomiatoids tend to nullify any pro- 
position that the enchodontoids are either ancestral or even related to the stomiatoids. 
Certain of these are as follows : 

1. In jaw structure there is a certain degree of similarity between the enchodon- 
toids and stomiatoids in that both possess a very strongly developed dentition. In 
the stomiatoids, however, the major tooth bearing bones are the premaxilla and 
maxilla, while the palato-pterygoid dentition is reduced or absent (Weitzman, 1967 : 
516, fig. 8). In the enchodontoids the palato-pterygoid dentition is strongly 
developed but that on the premaxilla and maxilla is reduced. 

2. The maxilla of the stomiatoids is the major component of the dermal upper 
jaw and supports supramaxillae (Weitzman, 1967 : 516, fig. 9). In enchodontoids 
the premaxilla is the major bone and the maxilla is variously reduced and never 
bears supramaxillae. 

3. There is considerable bone reduction in the stomiatoids, particularly in the 
neurocranium where the lateral prootic walls are lost in the anterior region of the 



ESPECIALLY MYCTOPHOIDS 195 

myodome (Greenwood, et al., 1966 : 372, fig. 5), but a large basisphenoid is retained. 
In enchodontoids the basisphenoid is never present and the prootic is complete. 

4. The shape of the ethmoid region in stomiatoids (Weitzman, 1967 : 514, 
fig. 7) is also quite unlike that of the enchodontoids. 

5. The peculiar specialization of the anterior vertebrae of the stomiatoids (Regan 
& Trewavas, 1930) also sets them apart from the enchodontoids. 

These five points provide sufficient evidence to assume that the two groups are 
not closely related. They do, however, share the features that ally them to the 
salmoniform complex within the Protacanthopterygii, for example, no fin-spines, 
abdominal pel vies, large number of branchiostegal rays, etc. A further feature 
which is shown by both groups is an adipose fin. This structure, or rather an 
impression of it in the matrix, is seen in a specimen of Enchodus gracilis from Senden- 
horst, Westphalia (Woodward, 1901 : 202). Impressions of an adipose fin are also 
seen in other specimens from Sendenhorst e.g. Sardinioides monasteri. 

The second major group with which the enchodontoids have been linked is the 
scopelids (Myctophiformes), and Woodward (1901) went as far as to suggest that they 
might furnish the ancestors of the myctophiforms. The hypothesis is untenable in 
the light of the host of specialized features in the enchodontoids which are more 
advanced than those shown by the basal myctophiform stock (e.g. Sardinioides and 
Aulopus). These specialized enchodontoid features are as follows : 

1. Basisphenoid and orbitosphenoid absent. 

2. Supraorbital and antorbital absent. 

3. Nasal bones absent. 

4. Supramaxillae absent. 

5. Interoperculum absent. 

6. No fulcral scales at the base of the tail. 

7. Incomplete squamation. 

8. No pelvic splint bone. 

These features which are absent or reduced in the enchodontoids are present in 
Aulopus, and all except the supraorbital are present in Sardinioides. Thus although 
the enchodontoids could not conceivably be ancestral to the Myctophiformes, they 
could still possibly be related to them. Again this latter proposition is negated by 
several factors : 

1. The maxilla enters the gape and is toothed in some genera of enchodontoids. 
In the Myctophiformes the maxilla never enters the gape and is never toothed. 

2. The premaxillary region in the enchodontoids might also set them apart since 
all of the Myctophiformes possess a true ascending process and an articular process. 
The enchodontoids do not appear to have a true ascending process, but have an 
enormously expanded articular process (Goody, 1968 : 228). 

3. The occurrence of the two groups in the fossil record also clearly separates 
them. At the base of the Cenomanian the Myctophiformes is represented by the 
primitive and generalized Sardinioides. The enchodontoids first occur in the Lower 
Cenomanian {Enchodus venator Arambourg, 1954 : 125) and are considerably more 
advanced in structure than the basal myctophiforms. 



ig6 UPPER CRETACEOUS TELEOSTS 

Presumably because of their similar habits, similar derivation from a salmoniform 
stock and similarity in evolutionary potential, many parallels can be drawn between 
the enchodontoids and the more advanced ' alepisauroid ' myctophiforms (Marshall, 

1955)- 

The first, and most striking, similarity between the enchodontoids and the 
advanced myctophiforms is the size and profusion of the buccal teeth. In both 
groups the dentition is concentrated on the palato-pterygoid, being absent from the 
vomer and only Enchodus showing a few scattered teeth on the endopterygoids. 
The structure of the palate, however, is somewhat different in the two groups. 
The palatine of the ' alepisauroids ' is extended and bears a row of large teeth, 
whereas the ectopterygoid remains short and often has no teeth (Text-figs. 78 
and 79). In the enchodontoids the palatine always bears a single terminal tooth 
and the ectopterygoid is more extensive and supports a row of large teeth. The 
end product is the same in that the palato-pterygoid is straight and bears a row of 
large ventral teeth. 

Great similarities are apparent in the structure of the dermal upper jaw, excluding 
whether the maxilla enters the gape or not. The fenestrated premaxillary pedicel in 
the enchodontoids is considered by Goody (1968 : 228) to be an enlarged articular 
process. A similar complex is seen in Lestidium and Anotopterus (Text-fig. 78) which 
both possess a true fenestrated ascending process, through which the slightly enlarged 
anterior dentary tooth projects much as in the enchodontoids. In Omosudis and 
Alepisaurus (Text-fig. 79) the premaxilla has a deep groove, rather than a discrete 
fenestra, between the narrow, attenuated ascending process and the expanded 
articular process. The maxillary head in the enchodontoids and the ' alepisauroids ' 
is similar in being simple. However that of the enchodontoids would appear to be 
primitive, whereas that of Alepisaurus is derived by simplification from a more 
complex pattern seen in the basal myctophiform stock. 

Some of the smaller skull bones which are present in the basal myctophiform stock 
are lost in the more advanced members and in this respect the advanced myctophi- 
forms resemble the enchodontoids more closely. For example, the nasals become 
fused with the modified mesethmoid in ' alepisauroids ' and in the enchodontoids the 
nasals are presumed to have disappeared completely. Supraorbitals and ant- 
orbitals are also absent in both groups. The basisphenoid and orbitosphenoid, 
which are both absent in the enchodontoids, are represented through all stages of 
reduction including total loss in the ' alepisauroids ' (Parr, 1929). Excluding the 
mesethmoid region, the neurocranial structure of Enchodus and Alepisaurus is 
somewhat similar as Goody (1968 : 229) has shown. 

In the Enchodontoidei an interoperculum is absent ; this peculiarity has also been 
observed in the suborder Cimolichthyoidei (p. 190). Several of the more advanced 
myctophiforms (Omosudis, Alepisaurus, Evermannella and Scopelarchus) have a much 
reduced interoperculum which is possibly in the process of being lost. It would 
appear that forms retaining a vertical suspensorium tend to have reduced inter- 
opercula, since presumably it no longer has a functional significance. 

Within the Enchodontidae (Palaeolycus, Text-fig. 43) there is a tendency for the 
body to become lengthened. This tendency is exhibited in the advanced ' alepi- 



ESPECIALLY MYCTOPHOIDS 197 

sauroids '. In both groups however the pelvics remain abdominal in position, and 
Marshall (1961 : 367) has indicated that the position of the pelvics in the myctophi- 
forms is correlated with the mode of life, rather than being correlated with locomotion 
as in the Acanthopterygii, the bathypelagic myctophiforms having in general 
abdominal pelvics, whilst the benthonic groups have sub-thoracic pelvics. This 
assumption substantiates one aspect concerned with the division of the enchodon- 
toids into pelagic and benthonic families since the Enchodontidae with abdominal 
pelvics were probably pelagic whilst the Eurypholidae with sub-thoracic pelvics 
were benthonic. 

The caudal skeleton is very similar to that of the cimolichthyoids. The tails 
of Enchodus (Text-fig. 42) and Eurypholis (Text-fig. 48) are remarkably alike and 
show features which duplicate those seen in Sardinioides (Text-fig. 72) and Aulopus 
(Text-fig. 93) among the Myctophiformes, and Prionolepis in the Cimolichthyoidei. 
These features include : 

1. The fusion of preural vertebra one with ural vertebra one, the compound 
centrum so produced supporting a stegural. 

2. The second ural vertebra present as a terminal half-centrum supporting four 
upper hypurals. 

3. Preural vertebra two has no neural spine, simply an expanded neural arch. 

The absence of basal fulcral scales above and below the caudal peduncle in the 
enchodontoids is as in Prionolepis. Both Sardinioides and Aulopus possess fulcral 
scales and are thus on a more primitive level in respect of caudal structure than the 
enchodontoids. 

The squamation of the Enchodontoidei is greatly reduced as it is in the ' alepi- 
sauroids '. This reduction is no doubt correlated with a distensible stomach (as in 
the Cimolichthyoidei, p. 191). 

In conclusion then, the suborder Enchodontoidei has two families of divergent 
habitat. The order Myctophiformes, likewise, is frequently divided into two main 
groups, and as Gosline, Marshall and Mead (1966 : 6) have pointed out the anatomical 
differences between the two groups may reflect differences in environment as much 
as in ancestry. These authors tentatively suggest that the Myctophiformes was 
originally a benthonic group and cite the presence of a sensory canal commissure 
across the frontals as evidence. In Eurypholis a median dorsal sensory canal is 
present behind the occiput and probably served the same function. The only other 
character which the Eurypholidae possess which is not present in the Enchodontidae 
is a roof to the post-temporal fossa. This fact would tend to indicate that the 
Eurypholidae is the more primitive of the two, but in most other respects the 
Eurypholidae show specializations suiting them to a benthonic existence. Thus the 
Enchodontoidei are best looked on as being derived probably from a pelagic ancestral 
stock within the Salmoniformes which evolved to produce a benthonic group as well 
as retaining its original pelagic habit. 



ig8 upper cretaceous teleosts 

(D) Suborder HALECOIDEI 

Three quite closely related genera, Halec, Phylactocephalus and Hemisaurida are 
included in this new suborder. There is a closer affinity between Halec and 
Phylactocephalus than between either of these and Hemisaurida. 

Halec has a relatively shallow head and body with certain advanced characteristics, 
particularly in respect of the neurocranium. These features are : 

1. Unroofed post-temporal fossa. 

2. Parietals separated in the mid-line. 

3. No basisphenoid. 

4. No orbitosphenoid. 

5. No supratemporal. 

6. Simplified trigemino-facialis chamber. 

All of these features are normally associated with advances encountered in the 
acanthopterygian grade (Patterson, 1964). The more primitive features seen in the 
genus are concerned with the body and are as follows : 

1. Pectoral girdle with mesocoracoid arch. 

2. Abdominal pelvic fins. 

3. Incompletely fused vertebral elements. 

4. Numerous intermuscular bones. 

The genus Phylactocephalus agrees with Halec in respect of the skull structure, but 
exhibits several more advanced features associated with the body. For example, 
the body is shortened (33 vertebrae, whereas Halec has 40 to 45), and the depth of 
both head and body are considerably greater than in Halec. Correlated with body 
shortening is the relative position of the fins, the dorsal having advanced and the 
pelvics having migrated to practically a sub-thoracic position. This migration may, 
however, merely be a further expression of the overall body shortening, rather than 
a major morphological, and presumably functional migration, since the pectorals 
still remain low down on the body. Also the pectoral girdle of Phylactocephalus may 
still retain the mesocoracoid arch present in Halec eupterygius (Text-fig. 62). 
Another character possibly associated with this general shortening and deepening of 
the body is the decrease in the number of branchiostegal rays (Phylactocephalus has 
10, whereas Halec has 15 or 16). 

Possibly these more advanced features which Phylactocephalus possesses indicate 
a different mode of life. Halec could have been benthonic and provided a stock from 
which a more active pelagic form such as Phylactocephalus was derived. Indications 
of a bottom-dwelling habit are given by the inflection of the lower edge of the 
mandible (already noticed in the probably benthonic Eurypholis, p. 193) and the 
high number of branchiostegal rays in Halec. In Phylactocephalus the number 
of rays is reduced to ten but the jaw still shows signs of being inflected. The 
overall body shape of Halec suggests certain recent benthonic forms such as 
Synodus. However Synodus has sub-thoracic pelvics which Marshall (1961 : 368) 
suggests are used together with the tail as props. Although Phylactocephalus has 
sub-thoracic pelvics, from its shortened, deepened, laterally compressed body it is 



ESPECIALLY M YCTOPHOIDS 199 

more likely to have been pelagic and not benthonic. The forward migration of the 
pelvics in Phylactocephalus would seem to have served a different function to that of 
Synodus, and may possibly be analogous in function to the advanced pelvics of the 
acanthopterygians. In these the forwardly placed pelvics cancel out the lift force 
produced by the pectorals when they are used as brakes (Harris 1938 : 37 ; 
Patterson, 1964 : 452). 

The structure of the jaws is identical in Halec (Text-fig. 61), Phylactocephalus and 
Hemisaurida (Text-fig. 67) providing the most obvious link between the three 
genera. The premaxillae are simple flattened plates with no articular or ascending 
processes. The oral border is toothed and forms half of the gape. The maxilla is 
thin, narrow and elongated, entering the gape behind the premaxilla. The head of 
the maxilla is little differentiated but does appear to be associated with a small 
palatine process. The teeth on the maxilla are of characteristic form, larger than 
those on the premaxilla and inclined forwards. A single, small ornamented supra- 
maxilla is associated with the posterior end of the maxilla. 

The dermal upper jaw is unspecialized and primitive but with certain advanced 
tendencies such as extension beneath the maxilla of the premaxilla, narrowing of the 
maxilla, and loss of one supramaxilla. 

Apart from the jaws the genus Hemisaurida (Text-figs. 67 and 68) shows several 
features not seen in either Halec or Phylactocephalus. The head and body remain 
shallow (as in Halec) but the snout has become elongated. Coupled with this 
elongation the jaw articulation has moved forwards to a point below the hind end 
of the orbit. Thus the suspensorium is inclined forwards. Ventrally the pre- 
operculum in Hemisaurida is produced into a prominent backwardly projecting 
spine (a similar spine has already been noticed in the ichthyotringoid Apateopholis 
Text-fig. 15, but in neither can any functional correlation be attached to it). The 
pectoral fin, like that of Halec and Phylactocephalus, is low on the flank, but is 
characterized by extension of certain of the rays. These extended rays are long 
and filiform, and the fin is not aliform as in the Exocoetidae or the Cretaceous 
Chirothricidae (Woodward, 1901 : 281, fig. 10). In both of these latter groups the 
pectorals are spread into wing-like structures and are positioned quite high on the 
flanks. The body of Hemisaurida is devoid of squamation, whereas both Halec 
and Phylactocephalus possess a complete covering of small scales. 

Although the caudal skeleton is not well represented in the three genera, sufficient 
can be made out to indicate that it is comparable {Halec, Text-fig. 64 and Hemisaurida, 
Text-fig. 69). The first preural vertebra is fused with the first ural vertebra and this 
compound centrum supports a stegural dorsally. The second ural centrum is much 
reduced and appears to support the four hypurals to the upper caudal lobe. The 
posterior haemal spines have a tendency to become laterally compressed. 

A feature of both Halec and Phylactocephalus, not seen in Hemisaurida, is the 
expanded nature of the bases of the anterior fin rays in all fins except the caudal. 
These expanded basal regions are ornamented. This feature is not normally 
associated with soft rays, but is occasionally found on the fin spines of Acantho- 
pterygii. 



200 UPPER CRETACEOUS TELEOSTS 

Little mention has been made in the past of the relationships and systematic 
position of the halecoids. Agassiz (1834) first used the generic name Halec, and also 
erected the family Halecidae, which he used to include the ' herring-like ' fishes, the 
clupeoids and salmonoids of Cuvier. Pictet (1850) and Davis (1887) retained the 
family, Davis adding the genus Phylactocephalus. Woodward (1901) retained the 
generic name Halec (including Phylactocephalus Davis in its synonymy), but put it 
in the family Enchodontidae, where it has remained to the present day (Bertin & 
Arambourg, 1958 ; Romer, 1966). 

Gregory (1933 : 204) mentioned the genus Halec in connection with the Enchodon- 
tidae, but pointed out that Halec was on a slightly lower level than Enchodus in that 
it possessed well developed maxillary teeth. 

Woodward (1901 : 270) also mentioned Hemisaurida as an imperfectly defined 
genus which perhaps belonged in the family Scopelidae. Romer (1966) followed 
Woodward, but definitely assigned Hemisaurida to the family Myctophidae. 

Direct relationship with the enchodontoids as constituted in this work can be 
dismissed because of the following features shown by the Halecidae : 

1. Elongate palatine with two rows of teeth. 

2. No fenestrated premaxillary pedicel. 

3. Supraorbital bone present. 

4. Supramaxilla present. 

5. Mesocoracoid arch present. 

Certain similarities seen in the two groups appear to be the result of convergent 
modes of life, and possibly similarities in evolutionary potential. These are as 
follows : 

1. A similar neurocranial structure in the loss and reduction of certain 

bones (orbitosphenoid, basisphenoid, etc.). 

2. Reduction in squamation. 

3. Trend towards body deepening and shortening. 

4. Trend towards snout elongation. 

5. Trend for forward migration of the pelvic fins. 

The Halecoidei are also in no way related to the Myctophiformes. The incomplete 
exclusion of the maxilla from the gape, and the absence of ascending and articular 
processes on the premaxilla completely exclude the halecoids from any connection 
with the myctophiforms. 

The halecoids do, however, show some points of similarity with the myctophiforms 
which are possibly correlated with convergent and/or parallel evolution, in space as 
well as time. The halecids occur throughout most of the Upper Cretaceous (Ceno- 
manian, Turonian and Senonian), and this occurrence coincides with the initial 
radiation of the myctophiforms. Both groups appear to have inhabited similar 
environments (occurring as fossils at the same horizons), and identical evolutionary 
forces presumably would have acted upon them. Further, since they were probably 
derived from a common ancestral stock within the Lower Cretaceous, it is not 
surprising that some similarities should exist. For example the caudal skeleton 



ESPECIALLY MYCTOPHOIDS 201 

(Text-fig. 64) resembles that of the basal myctophiforms (Aulopus, Text-fig. 93 and 
Sardinioides, Text-fig. 72) in the fusion of preural vertebra one and ural vertebra 
one. It has already been shown elsewhere that the same state of affairs exists in 
both the Cimolichthyoidei and the Enchodontoidei. This character then is of no 
real value in the understanding of relationships. It would appear that the consolida- 
tion of the caudal skeleton is general throughout the more basal Protacantho- 
pterygii, and the first step is the fusion of the first preural vertebra with the first 
ural vertebra. 

The nature of the neural components of preural vertebra two (the first free preural 
vertebra) is of interest. In the halecids the neural spine is complete in Hemisaurida 
but reduced in Halec. In the myctophiforms the neural spine is absent. This 
reduction and loss of the neural spine is probably an expression of the further 
reduction and consolidation of the caudal skeleton. The basal salmoniforms 
retain the neural spine on preural vertebra two and in this respect the halecoids 
are closer to the basal stock. 

The Halecoidei as a whole present a constellation of primitive and advanced 
characters. A similar state of affairs has already been shown in both the Cimo- 
lichthyoidei and the Enchodontoidei. These have been placed on a structural grade 
below that of the myctophiforms and acanthopterygians, due to several primitive 
characters. The evolutionary trends shown by the halecids agree with those already 
seen in the cimolichthyoids and enchodontoids and with those in the myctophiforms. 
These trends are those which align the group within the Protacanthopterygii (sensu 
Greenwood, et al., 1966), and can be summarized as follows : 

1. Tendency for premaxilla to exclude maxilla from the oral border. 

2. Tendency for supraorbital to be lost. 

3. Tendency for supramaxillae to be lost. 

4. Reduction in the roof of the post-temporal fossa. 

5. Tendency for pel vies to advance. 

Tendency for the caudal skeleton to become consolidated, the first step 
being the fusion of preural vertebra one with ural vertebra one. 

Tendency for the body to become shortened and deepened. 

Certain basic features, as opposed to trends, suggest that the halecoids are closest 
to the salmoniform grade at the base of the Protacanthopterygii. These would 
include the retention of a mesocoracoid arch ; absence of an ascending process on 
the premaxilla ; toothed maxilla ; supraorbital present ; numerous branchio- 
stegal rays ; finally the large number of pelvic rays. Following on from these 
considerations it is apparent that the halecoids are not directly related to any of the 
above mentioned groups (enchodontoids, cimolichthyoids, myctophiforms or 
acanthopterygians), but are related only in so far as they are contemporary descend- 
ants from an earlier stock within the Salmoniformes. 



202 UPPER CRETACEOUS TELEOSTS 

Order MYCTOPHIFORMES 

The ordinal ranking given to this group is based on the fact that it can be separated 
from the Salmoniformes and the Ctenothrissiformes at least by the opening of the 
Upper Cretaceous. This is in agreement with the ordinal rank given by Gosline, 
Marshall and Mead (1966 : 5) although, as these authors point out, there is apparently 
no one feature that will separate all myctophiforms from all salmoniforms. Green- 
wood, et al. (1966 : 366), using the same criterion, that they could find no diagnostic 
characters which would satisfactorily distinguish these forms from those they group 
as the Salmoniformes, therefore consider the Myctophiformes to be merely a suborder 
within the order Salmoniformes. This ordinal separation will be considered in 
greater detail further on in this discussion. 

The term Myctophiformes is considered preferable to the alternatives previously 
used, Iniomi and Scopeliformes, since the group has normally been collectively 
termed the myctophoids. 

A re-evaluation of Woodward's (1901 : 235) family Scopelidae, in particular the 
Cretaceous genera, has shown that several of these do not belong in the Myctophi- 
formes. 

Sedenhorstia (Microcoelia) White and Moy-Thomas (1941) and Dactylopogon von 
der Marck (1858) are both elopiforms (Goody, 1969). 

Ichthyotringa (Rhinellus) Cope (1878), and Apateodus Woodward (1901), are both 
primitive salmoniforms (see pp. 169 to 176). 

The remaining Cretaceous genera are more nearly referable to the myctophiforms : 
Sardinioides von der Marck (1858), Cassandra (Leptosomus) White and Moy-Thomas 
(1940), Acrognathus Agassiz (1843), Nematonotus Woodward (1901), and Sardinius 
von der Marck (1858). 

Sardinioides is indisputably a myctophiform and Cassandra would appear to be 
synonymous with it. The species of Sardinioides considered in the systematic 
account {Sardinioides minimus) was originally described as Cassandra minima. 
This species represents a form very close to the ancestral myctophiform stock. 
Sardinioides is a genus which has received some attention in the past as being a form 
from which both the present day Myctophiformes and also possibly the Acantho- 
pterygii could have been derived. Patterson (1964) considered the latter possibility 
and examined both Sardinioides and Aulopus as possible acanthopterygian ancestors. 
He concluded that they were not on the direct line of descent of the acantho- 
pterygians, but that this ancestry was closer to the Ctenothrissiformes. 

A table is given below of the major osteological characters shown by the Sal- 
monoidei (the most primitive group within the Salmoniformes) as illustrated by 
Salmo and the corresponding characters of the Myctophiformes (from Sardinioides) 
and the Ctenothrissiformes (from Patter sonichthys) . Similarity to Salmo is indicated 
by a cross ( x ) . 



ESPECIALLY MYCTOPHOIDS 




Table i 






Myctophiformes 




Ctenothrissiformes 


Sardinioides 




Patter sonichthys 


X 




X 


X 




X 


X 




X 


X 




X 


Prominent ascending process 


Small 


ascending process 


Prominent articular process 


Small articular process 


Very long toothed premaxilla 




X 


Large, untoothed maxilla 




X 


excluded from gape 






2 reduced supramaxillae 




X 


X 




X 



203 



Salmoniformes 
Salmo 
Parietals meet in mid-line 
Post-temporal fossa roofed 
Orbitosphenoid present 
Basisphenoid present 
No ascending premaxillary 

process 
Articular premaxillary 

process present 
Small, toothed premaxilla 
Large, toothed maxilla 

entering the gape 

2 large supramaxillae 
Teeth on vomer, palatine, 

endopterygoid, 

ectopterygoid 
Supraorbital present 
Antorbital present 
Expanded posterior 

infraorbitals 
Numerous branchiostegals 
Mesocoracoid arch present 
Pectorals low down, 18 rays 

3 postcleithra 

Numerous accessory radials 
Pelvics abdominal, 13 rays 
Pelvic splint bone present 
Imcompletely fused vertebral 

elements 
Numerous epineurals 
No fin spines 
Adipose fin present 
Stegural present 
2 free ural vertebrae 
17 branched caudal rays 
Fulcral scales absent 
Cycloid scales on body 
6 hypurals 

The differences noticed between the three groups are all relatively minor ones, and 
only one factor in the whole list would exclude a direct descent of the two groups 
from the salmoniforms. This factor is the absence of basal fulcral scales at the base 
of the caudal fin in Salmo. This however is a primitive feature left over from some 
halecostome ancestor and is lost in later myctophiforms and ctenothrissiforms. 
Likewise it has presumably been lost in the Recent salmonoids. 

The list illustrates the strong basal convergence of three groups which are separated 
ordinally. It thus would appear that the sardinioid group and the ctenothrissoid 
group, at the beginning of the Upper Cretaceous, were evolving along similar lines, 
both having been derived from a common ancestral stock within the salmoniforms. 



Supraorbital absent 

X 
Unexpanded infraorbitals 

9 (numerous in Aulopus) 
Mesocoracoid absent 
Pectorals on flank, 13 rays 

2 postcleithra 

3 or 4 accessory radials 
Pelvics abdominal, 9 rays 

X 
X 



Supraorbital absent 

X 
Unexpanded infraorbitals 

8 or 9 branchiostegals 
Mesocoracoid absent 
Pectorals on flank, 13 rays 

2 postcleithra 

3 or 4 accessory radials 
Pelvics sub-thoracic, 8-9 rays 

X 
X 



X 
X 
X 
X 
One free ural vertebra 


X 
2 spinous rays in dorsal 
Adipose fin absent 

X 
One free ural vertebra 


X 
Fulcral scales present 
Cycloid/ctenoid scales 

X 


X 
Fulcral scales present 
X 

X 



j.. I UPPER CRETACEOUS TELEOSTS 

From the above considerations it can be seen that any division between the three 
will be somewhat arbitrary. One feature, the presence of a definite ascending 
process on the premaxilla, separates the myctophiforms and ctenothrissiforms from 
the salmoniforms. However the ancestral forms from which these arose presumably 
must have had at least the forerunner of an ascending process. Points for and 
against ordinal separation can be argued, but in view of the potential within each 
group, especially the Ctenothrissiformes, it is convenient to give both groups ordinal 
separation. 

Turning to a consideration of the myctophiforms alone, a direct comparison can be 
made between the Cenomanian Sardinioides and the recent genus Aulopus. This 
latter genus has often been cited in phylogenetic considerations of teleosts (Patterson, 
1964), and the skull has been figured by both Woodward, 1902 and Regan, 1911. 
A complete osteological description has been compiled for comparative purposes 
and is set out below. 



Aulopus filamentosus (Bloch) 
(Text-figs. 86-93) 

Neurocranium. The neurocranium is shown in dorsal, ventral and lateral views 
in Text-figures 86, 87 and 88. In general shape the neurocranium is long and broad. 
The skull-roof is flattened but inclines ventrally in the postero-lateral region. The 
neurocranium attains its greatest width posteriorly in the region of the pterotics. 
Much of the skull-roof is formed from the large frontals which are somewhat con- 
stricted between the orbits but widen posteriorly. The surface of the frontal is 
smooth although some surface markings in the form of feeble ridges occur above the 
curved orbital margin. The sensory canals run within the bone and their course is 
indicated externally by broad ridges. The frontals meet in a more or less straight 
suture in the mid-line. Postero-laterally the frontal contacts the sphenotic and 
pterotic and posteriorly the broad parietal. The tube which contained the supra- 
orbital sensory canal opens postero-laterally above the sphenotic and posteriorly just 
in front of the parietal. A further posterior branch extends back into the parietal. 
The tube passes anteriorly, where it receives a further tube from the medial region 
of the frontal. This medial connection is continuous across the skull-roof (the 
frontal commissure of Gosline, Marshall & Mead, 1966 : 7). Weitzman (1967 : 
527) points out that this cross-connection is a common occurence in the myctophi- 
forms, but is rare among the salmoniforms. Anteriorly the frontal meets the 
rear end of the mesethmoid, and just posterior to this the supraorbital sensory 
canal opens on to the nasal. A prominent supraorbital bone, at the anterior end of 
the orbit, is attached to the lateral edge of the frontal. 

The lateral ethmoid extends from beneath both the frontal and the supraorbital 
forming the anterior limit of the orbit. It is large and well ossified and attached 
mainly to the dorsal surface of the parasphenoid. Ventro-laterally the lateral 
ethmoid is free from the parasphenoid and has a cartilaginous connection with 
the postero-dorsal region of the palatine. In the mid-line the lateral ethmoids 



ESPECIALLY MYCTOPHOIDS 



205 



almost meet one another, and the median laminar orbitosphenoid behind. The edge 
of the lateral ethmoid has a definite facet for the articulation of the lachrymal. 

The nasals are small and elongate with rolled edges resting on the dorsal surfaces 
of the frontal and mesethmoid. The terminal part of the supraorbital sensory canal 
passed within a groove on the dorsal surface of the nasal. 



vo 



so.sc 




infs.c 



Fig. 86. Aulopus filamentosus . Neurocranium in dorsal view. The broken line on 
the left side of the figure indicates the course of the sensory canals. 



206 UPPER CRETACEOUS TELEOSTS 

The mesethmoid is prominent, extremely well ossified and with its maximum 
width equalling the interorbital width of the frontals. The mesethmoid is blunt 
posteriorly where it contacts the frontals, and is hollowed out internally where it is 
filled with cartilage. Weitzman (1967 : 526) states that Aulopus possesses a 
myodome bone in this region, separate from the ethmoid ossification, also that 
capsular ethmoid bones appear to be fused to the mesethmoid. The lateral pro- 
jection of the mesethmoid inclines ventrally to terminate in an articulatory facet. 
These wings are associated with corresponding, but smaller, projections from the 
vomer and together provide the articulatory facet for the anterior region of the 
palatine. Anteriorly the mesethmoid edge has a small median U-shaped indentation 
which houses a small circular knob of cartilage, presumably the rostral cartilage. 
Arising from the centre of the bone and passing antero-laterally are two smooth 
ridges which extend right to the edge. The anterior edge of the mesethmoid 
between these two ridges provides the articulatory region for the short ascending 
processes of the premaxilla. 

The vomer is attached anteriorly to the ventral surface of the mesethmoid and its 
internal extensions parallel those of the mesethmoid. Dorso-laterally a facet 
between the dorsal surface of the vomer and the ventral surface of the mesethmoid 
contains the articular head of the maxilla. Antero-ventrally the vomer has two 
oval patches of teeth, on either side of the mid-line, each containing seven or eight 
small, acutely pointed, recurved teeth. The anterior end of the palatine lies against 
the lateral vomerine region so that the palatine teeth are continuous with those on 
the vomer. The vomer is produced posteriorly into a short shaft applied to the 
ventral surface of the parasphenoid. 

The parietals are transversely elongated bones meeting in the mid-line, contacting 
the frontals anteriorly, the pterotics laterally and the supraoccipital and epiotics 
posteriorly. The dorsal surface of the parietal, like the frontal, is unornamented 
except for a low transverse wavy ridge extending from the lateral edge almost to the 
mid-line. Laterally, the transverse portion of the supratemporal lies on the dorsal 
surface of the parietal, fitting against the posterior edge of this ridge. More medially 
the supratemporal sensory canal passed across the parietal behind the ridge. The 
supraorbital sensory canal terminates posteriorly on the parietal. 

The supraoccipital appears on both dorsal and posterior faces of the skull. The 
small posteriorly directed supraoccipital crest does not extend above the plane of the 
skull-roof. 

The pterotic forms the postero-lateral region of the neurocranium, meeting the 
sphenotic anteriorly and the frontal and parietal medially in the roof of the post- 
temporal fossa. A slight crest on the pterotic passes postero-laterally. This crest 
contains the otic branch of the infraorbital sensory canal which opens midway along 
the crest, from where it continues in a flanged groove. Postero-medially the supra- 
temporal is closely applied to the pterotic. The extreme posterior end of the 
pterotic is drawn out into a small acute spine which lies ventral to the longitudinal 
limb of the supratemporal. The lateral line canal passed through the supratemporal 
to enter the pterotic posteriorly. The posterior pore in the pterotic for this sensory 



ESPECIALLY MYCTOPHOIDS 



207 



canal also transmitted the preopercular sensory canal. Laterally the pterotic is 
deeply excavated as the dilatator fossa. The anterior part of this fossa is however 
formed from a postero-dorsal excavation of the sphenotic, and the suture between 
the pterotic and sphenotic is visible both above and below the fossa. 

The post-temporal fossa is large and partially roofed, the roof being composed 
anteriorly of frontal and posteriorly of parietal and pterotic. The lateral wall and 




spo 



f.hm 



VII hm 



par 



10mm 



Fig. 87. Aulopus filamentosus. Neurocranium in ventral view. 



208 



UPPER CRETACEOUS TELEOSTS 



part of the floor are formed of pterotic whilst the medial wall and remainder of the 
floor consist of frontal, parietal and epiotic. Gosline, Marshall and Mead (1966 : 5) 
used the presence of a post-temporal fossa in Aulopus as evidence for an elopoid 
derivation of the myctophiforms, believing a post-temporal fossa to be absent in 
the salmoniforms. As Weitzman (1967 : 528) has indicated this is an erroneous 
interpretation of the nature of a post-temporal fossa, and post-temporal fossae 
are present in most salmoniforms. It has been shown in the fossil salmoniform 
suborders considered in this paper, that the ichthyotringoids and the enchodontoids 
have representatives with completely roofed fossae, while the cimolichthyoids 
have a partially roofed fossa. In the floor of the fossa, near to the posterior edge, 
is a transversely orientated oval fenestra filled with cartilage. 

The epiotic joins the supraoccipital medially both on the roof of the skull and on 
the posterior face. Medial to the opening of the post-temporal fossa the epiotic has 
a prominent dorsal knob which provides the articulatory area for the dorsal limb of 
the post-temporal. 




f.o.a 



Fig. 



Aulopus filamentosus. Neurocranium in lateral view. 



The sphenotic forms the postero-dorsal region of the orbit, contacting the frontal 
dorsally and the pterotic posteriorly. The sphenotic is partially obscured dorsally 
by the last infraorbital. This infraorbital is intimately connected to the edges of 
both the sphenotic and frontal. This intimate connection of the last infraorbital 
with sphenotic and frontal was noticed in Sardinioides (p. 154), and is also seen in 
Nematonotus (Text-fig. 94). 

The hyomandibular facet is formed of sphenotic, pterotic and prootic. The upper 
margin of the facet also forms the ventral border of the dilatator fossa. The facet is 
in two distinct regions, an anterior cup-shaped depression formed mainly of sphenotic 
(with a ventral inclusion from the prootic), and a posterior elongated oval cup within 
the pterotic. 



ESPECIALLY M YCTOPHOIDS 209 

The prootic forms much of the lateral wall of the cranial cavity and myodome as 
well as a certain part of the posterior wall of the orbit. Ventrally the prootic is thin, 
meeting the parasphenoid below and the exoccipital behind. More dorsally the 
prootic divides into two layers, an outer layer inclining dorso-laterally and an inner 
layer passing medially to form the roof of the posterior myodome. The outer layer 
contacts the sphenotic in the hyomandibular facet and the pterotic below the mid- 
region of the same facet. 

The lateral region of the cranial cavity below the hyomandibular facet takes the 
form of a triangular depression. The posterior boundary of the lateral wall of the 
cranial cavity is formed from a stout strut of bone composed dorsally of pterotic and 
ventrally of exoccipital. The intercalar is superficial in position and plastered 
around this strut, covering the suture between the pterotic and exoccipital. The 
intercalar is, then, an elongated oval bone appearing on both the lateral and posterior 
faces of the neurocranium. 

The inner sheet of prootic contacts its partner in the mid-line to form the prootic 
bridge separating the myodome from the cranial cavity. The arms of the basi- 
sphenoid articulate with the anterior edge of the prootic bridge. The trigemino- 
facialis chamber is divided by a bony wall into a medial pars ganglionaris and a 
lateral pars jugularis. The pars jugularis is a short horizontal canal, narrow 
posteriorly, and lying lateral to the prootic bridge in the angle between the internal 
and external layers of the prootic. There are two foramina entering the pars 
jugularis from the pars ganglionaris, arranged one above the other. The ventral 
foramen is the facial foramen through which passed the hyomandibular branch of 
the facial nerve. The dorsal foramen is the trigeminal foramen and through this 
passed the trigeminal, profundus and superficial ophthalmic nerves. The pars 
jugularis has four external openings ; the posterior one is small, merely transmitting 
the jugular vein. Dorso-laterally there is an extremely large, elongate foramen 
reaching up almost to the hyomandibular facet. This foramen directed the hyo- 
mandibular nerve upwards and outwards to the hyomandibular. The anterior 
opening of the pars jugularis is large and triangular, in the hind wall of the orbit. 
Through it the jugular vein, the orbital artery, the trigeminal, profundus and super- 
ficial ophthalmic nerves passed. The orbital artery entered the pars jugularis 
through the fourth opening which is ventro-lateral in position and close to the 
posterior opening of the pars jugularis. 

The parasphenoid is long and straight, extending for practically the entire length 
of the neurocranium. The anterior end, lying on the dorsal surface of the vomer, is 
narrow. Below the lateral ethmoids the parasphenoid is considerably widened, due 
to the presence of large lateral flanges, inclined slightly ventrally. In the mid-dorsal 
line of the parasphenoid a thickened pad of bone between the ascending processes 
provides the attachment region for the basisphenoid pedicel. The ascending 
processes arise at the level of the basisphenoid pedicel but are not particularly deep 
and contact the prootics dorsally in the wall of the myodome. Near the base of the 
ascending process a slight depression marks the articulatory point of the first 
pharyngobranchial. Above the depression, the foramen of the internal carotid 



210 UPPER CRETACEOUS TELEOSTS 

artery is visible. Posteriorly the parasphenoid articulates with the ventro-lateral 
edges of the basioccipital, ending posteriorly in two short processes below the hind 
end of the basioccipital. 

The basioccipital, posteriorly, forms the lower part of the occipital condyle and 
laterally it contacts the exoccipital above and the prootic in front. At the rear end 
of the lateral face of the basioccipital a pitted depression marks the point of insertion 
of supracleithral ligaments. Internally the basioccipital forms the posterior walls 
of the otolith chambers, and also separates them medially. The ventral surface of 
the basioccipital is deeply concave where it forms the roof to the myodome. The 
myodome is a deep cone-shaped recess ending blindly at the posterior end of the 
basioccipital. Ventrally, below the trigemino-facialis chamber, the anterior end of 
the otolith chamber is formed of prootic and is separated from the myodome by a 
membranous sheet. The otolith chamber is continued anteriorly as a recess on the 
internal face of the prootic. 

The exoccipitals contact each other in the mid-line of the posterior face of the 
skull, enclosing the foramen magnum completely. Ventrally each exoccipital con- 
tributes to the formation of the occipital condyle. Ventro-laterally the exoccipital 
is perforated by a large foramen through which the vagus nerve passes. Antero- 
ventrally the exoccipital contributes to the formation of the wall of the otolith 
chamber. The exoccipital, near its anterior edge, has a small foramen which 
transmits the glossopharyngeal nerve. The intercalar is entirely superficial in 
position, taking no part in the formation of the cranial wall itself. It is wrapped 
around the postero-ventral corner of the neurocranium beneath the opening of the 
post-temporal fossa. The intercalar is produced into a thickened knob of bone for 
the articulation of the ventral limb of the post-temporal. 

Ventrally within the orbit the basisphenoid is Y-shaped with a long, incompletely 
ossified pedicel, and shorter well ossified dorsal arms. These arms attach to the 
medial edges of the prootic and the anterior edge of the prootic bridge. The orbital 
face of the prootic contains the large anterior opening of the pars jugularis. Medial 
to this opening a further small foramen is visible which transmits the oculomotor 
nerve. Dorso-laterally the prootic meets the sphenotic and dorso-medially the 
pleurosphenoid. The pleurosphenoids are separated in the mid-line by a small 
optic fenestra, bordered ventrally only by the basisphenoid. A slight groove 
extends from the anterior opening of the pars jugularis to pass dorsally across the 
face of the pleurosphenoid. This groove indicates the course of the superficial 
ophthalmic nerves towards the ventral surface of the frontal. 

The orbitosphenoid is large, flattened and in the anterior orbital region. Ventrally 
it almost meets the parasphenoid and anteriorly the lateral ethmoids. Posteriorly 
the orbitosphenoid contacts the pleurosphenoids. 

Circumorbital series. There are eight circumorbital bones, of which seven are 
associated with the infraorbital sensory canal. The eighth is the supraorbital 
which is not canal bearing. The antorbital is a small fragile slip of bone contacting 
the supraorbital postero-dorsally, and the lachrymal antero-ventrally. The lach- 
rymal is the largest bone in the series and overlaps the upper regions of the maxilla. 



ESPECIALLY MYCTOPHOIDS 211 

The infraorbital sensory canal runs within a tube in the lachrymal and an antero- 
dorsal opening passed the sensory canal to the antorbital. The lachrymal extends 
back to the centre of the orbit where it becomes shallower. The remainder of 
the circumorbital ring is composed of five unexpanded infraorbitals, each bearing a 
flanged groove for the infraorbital sensory canal. 

Hyopalatine bones. The hyopalatine bones are shown in medial view in Text- 
figure 89. The hyomandibular is neither deep nor broad. The head of the bone is 
double and the two condyles are connected by a thin lamella of bone. The opercular 
process is large and extends postero-ventrally. The lateral face of the hyomandi- 
bular has a shallow crest arising from the anterior head region and curving postero- 
ventrally to run along the posterior edge of the bone. This crest disappears near 
the ventral end of the hyomandibular. The upper region of the preoperculum lies 
against the posterior edge of this crest. Ventrally the hyomandibular tapers and is 
terminated by cartilage. The anterior region is thin and overlain ventrally by the 
postero-dorsal edge of the metapterygoid. The hyomandibular branch of the facial 
nerve enters the bone through a small foramen close to the anterior region of the head. 
Within the bone the nerve divides and the opercular nerve passes through a foramen 
on the posterior edge of the bone while the hyoidean and mandibular branches leave 
through a larger foramen in the angle between the opercular process and the crest 
on the lateral face of the bone. 

The symplectic is long, thin and narrow and forwardly inclined. Its connection 
with the hyomandibular is cartilaginous and the interhyal also contacts this cartilage 



hm.h hmVII 



op. p. hm 



mx.p.p 



sop 




ang 



Fig. 89. Aulopus filamentosus. Hyopalatine and opercular bones and mandible of 

the left side in medial view. 



212 UPPER CRETACEOUS TELEOSTS 

block. The symplectic is acutely pointed ventrally and lies in a medial groove on 
the quadrate. 

The quadrate is large and triangular with a transverse condyle. The bone is thin 
but its anterior and posterior edges are slightly thickened. The symplectic groove is 
long and shallow, ending ventrally just behind the condyle. The posterior thickened 
edge is grooved slightly for the leading edge of the preoperculum. The anterior edge 
of the quadrate is sutured to the ectopterygoid. 

The ectopterygoid is bent through a slight angle. Anteriorly it meets the palatine 
in an elongate oblique suture, and in its anterior half bears an elongate, ventral tooth 
patch. The individual teeth are minute, especially posteriorly. The dorsal edge of 
the ectopterygoid is grooved for the reception of the endopterygoid. 

The metapterygoid is a well ossified, thin sheet of bone contacting the quadrate 
ventrally. More dorsally the metapterygoid has a prominent internal horizontal 
flange widening anteriorly. The posterior region of the endopterygoid is inserted 
on to the ventral surface of the anterior part of this flange. Postero-dorsally the 
metapterygoid overlies the antero-ventral part of the lateral face of the hyoman- 
dibular. 

The endopterygoid is thin and horizontal. The dorsal surface is slightly concave 
and articulates medially with the ventral surface of the lateral parasphenoid flange. 
The endopterygoid bears an oval patch of teeth ventrally. As on the ectopterygoid 
the teeth are minute and give the appearance of being nothing more than a roughened 
surface. 

The palatine is elongate and thicker than the ectopterygoid, inclining, anteriorly, 
towards the mid-line. Postero-dorsally the palatine is thickened medially where it 
connects with the ventro-lateral region of the lateral ethmoid. The maxillary 
process of the palatine is hollow and filled with cartilage. Gosline, Marshall and 
Mead (1966 : 2) have suggested that the palatine bone, in respect of its articulation 
with the maxilla, is different from the salmoniforms. It certainly is significantly 
different from that of the elopoids (Gosline, Marshall and Mead, 1966, fig. 1 ; Goody, 
1969), but Weitzman (1967 : 528) points out that many salmoniforms have the 
same arrangement as in the myctophiforms. On the medial face of the maxillary 
process the lateral wing of the mesethmoid attaches. Below the maxillary process 
is a further facet for the reception of the lateral vomerine wing. Ventrally the 
palatine has a broad band of teeth. The teeth are of two main types, minute 
marginal teeth continuous with those on the ectopterygoid, and larger, acutely 
pointed teeth in two rows. The tooth bearing region tapers posteriorly so that at 
the extreme posterior end only the internal row of palatine teeth remains. 

Dermal upper jaw. The dermal upper jaw is shown in lateral view in Text-figure 
90. The premaxilla is long and narrow for most of its length, forming the entire oral 
border of the mouth. Anteriorly a short broad ascending process meets its partner 
in the mid-line and both rest on the anterior edge of the mesethmoid. The internal 
face of the ascending process is hollowed out and rests on cartilage above the 
vomerine region. From the postero-lateral edge of the ascending process there is a 
short, stout articular process. The premaxilla is furnished with minute marginal 



ESPECIALLY MYCTOPHOIDS 



213 



teeth. Anteriorly the teeth are arranged into three distinct rows, a minute row of 
marginal teeth, a median row of slightly larger teeth and an internal row of much 
larger teeth. All of the teeth are similar, being long, thin, needle-like and slightly 
recurved. Posteriorly the teeth decrease in size and are confined to two rather than 
three rows. 

The anterior part of the maxilla, like that of the premaxilla, is inclined medially. 
The dorsal surface of the anterior part of the bone is excavated as a shallow depres- 
sion into which the maxillary process of the palatine fits. The actual head of the 
maxilla is small and rounded, sandwiched between the dorsal surface of the vomer 
and the ventral surface of the antero-lateral mesethmoid projection. The articular 
process of the premaxilla fits against a shallow concave facet on the anterior edge of 
the maxilla just lateral to the maxillary head. The maxilla is an elongate shaft of 
bone, stout anteriorly but becoming flattened and deepened posteriorly. The lateral 
face is smooth and dorsally supports two supramaxillae. Both supramaxillae are 
small and elongate, the anterior one being the smaller. 



epo pa d.f 



d.pcl 




smx2 
iop pop ang 

Fig. 90. Aulopus filamentosus. Skull in lateral view. 



10mm 



Mandible. The mandible is shown in medial and lateral views in Text-figures 89 
and 90. It is long and deepened posteriorly. The dentary forms the greater part 
of the ventral border, the whole of the oral border and a considerable proportion of 
the lateral face. In lateral view the dentary is V-shaped with the indentation 
occluded by the anterior part of the articular. The symphysis is constricted and 
blunt and each mandible has an inflected ventral edge. Ventro-laterally a tube 
within the bone opens to the exterior by a row of small pores. The mandibular 
sensory canal runs within the tube. Posteriorly the ventral limb of the dentary 



214 UPPER CRETACEOUS TELEOSTS 

meets the articular in a straight suture. The upper limb of the dentary forms the 
tooth bearing margin of the jaw, the teeth being borne on a stout medially directed 
flange. The teeth are present in three major rows and are similar to those on the 
upper jaw. 

The articular forms the posterior jaw region and has a shallow, concave, trans- 
versely arranged facet. The articular is produced behind the facet into a small 
retroarticular process. The mandibular sensory canal runs within the articular and 
opens on to the surface through several pores. The dorsal part of the articular is thin 
although the posterior edge is thickened. 

The angular is a small knob of bone applied to the extreme postero-ventral surface 
of the articular. 

Opercular bones. The opercular bones are shown in medial and lateral view in 
Text-figures 89 and go. The operculum is reduced in extent whereas the sub- 
operculum is greatly expanded. The operculum has straight anterior and dorsal 
edges, but is rounded posteriorly. The lateral face of the bone is completely smooth. 
The opercular facet is prominent, and near to the dorsal end of the anterior edge of 
the operculum. The facet is supported by an oblique flange crossing the internal 
face of the operculum. 

The suboperculum is greatly expanded forming the entire posterior edge of the 
opercular cover. Its dorsal border lies medial to the ventral and posterior edges of 
the operculum, and a large pointed process from the antero-dorsal corner extends 
upwards medial to the operculum. The anterior edge of the suboperculum is 
thickened and overlaps the posterior edge of the interoperculum medially. 

The interoperculum is roughly triangular with a vertical posterior edge. It 
ends anteriorly just behind the angular region of the mandible. 

The preoperculum is large and curved forwards ventrally. The dorsal termination 
of the preoperculum lies immediately above the opercular process. Dorsally the 
leading edge rests against the hyomandibular crest, but below the crest the edge of 
the preoperculum expands anteriorly to overlap the lateral face of the ventral end 
of the hyomandibular. Ventrally the preoperculum lies against the posterior edge 
of the quadrate. The preopercular sensory canal runs within a tube in the centre 
of the lateral face of the preoperculum, and gives off large posteriorly directed pores 
for subsidiary branches of the canal. The extreme posterior edge of the preoper- 
culum lies over the anterior edges of both the operculum and suboperculum, and 
covers much of the interoperculum. On the ventral part of the preoperculum the 
sensory canal tube opens by three large pores. The anteriormost opening is directly 
behind the retroarticular process of the mandible. 

Hyoid arch and branchiostegal rays. The hyoid arch consists anteriorly of a large 
ventral hypohyal and a smaller dorsal hypohyal connected to it by cartilage. The 
ceratohyal is long and shallow, whereas the epihyal is shorter and triangular. A 
small mesially constricted interhyal connects the epihyal to the interspace of cartilage 
at the base of the hyomandibular. Both the ceratohyal and the epihyal have a 
continuous groove on their lateral faces for the hyoidean artery. A median, 
vertically arranged urohyal extends posteriorly from its connection with the ventral 



ESPECIALLY MYCTOPHOIDS 



215 



hypohyals. Fifteen branchiostegal rays occur on the hyoid arch, all being of similar 
size and shape. Eight are present on the ceratohyal and 7 on the epihyal. 

Branchial arches. The basibranchials are fused into a single median elongate 
element, supporting three pairs of hypobranchials. Five discrete ceratobranchials 
are present, the first two of which bear toothed gill-rakers. Gill-rakers are absent on 
the third and fourth ceratobranchials but teeth are still present. The fifth cerato- 
branchial is the only remaining ossification of the fifth arch. It is enlarged into a 
prominent tooth plate, the inferior pharyngeal bone. Weitzman (1967 : 530) finds a 
small cartilaginous fifth epibranchial, ligamentously attached to the fifth cerato- 
branchial. Four epibranchials are present and, like the ceratobranchials, the first 
two bear toothed gill-rakers but the third and fourth simply bear teeth. There are 
four pharyngobranchials ; the first is small and attaches to the parasphenoid. 
Pharyngobranchials 2, 3 and 4 are fused into a tooth plate which opposes the inferior 
pharyngeal bone. Weitzman (1967 : 530) reports that a fifth pharyngeal tooth 
plate occurs in Aulopus, but he goes on to suggest that by its morphological relation- 
ships this tooth plate may actually belong to the fourth arch. Nelson (1967 : 286) 
indicates that Aulopus has a fifth tooth plate as a separate element. 

Pectoral girdle and fin. The pectoral girdle is shown in medial view in Text-figure 
91. The supratemporal is composed of transverse and longitudinal limbs. The 
transverse limb lies on the postero-dorsal regions of both the pterotic and parietal 
forming a part of the roof of the post-temporal fossa (Text-fig. 86) . This extension 
contains the supratemporal sensory canal. The longitudinal limb is the shorter of 



d.pcl 



sea 



v. pel 




Fig. 91. Aulopus filamentosus. Pectoral girdle of the left side in medial view. 



216 UPPER CRETACEOUS TELEOSTS 

the two. The posterior part of the longitudinal limb receives the lateral line sensory 
canal from the post-temporal. The sensory canal divides within the supratemporal 
and the supratemporal commissure passes medially whilst the main branch extends 
anteriorly. The post-temporal has an elongated strut-like dorsal limb articulating 
with the dorsal surface of the epiotic, and a shorter, unexpanded ventral limb articula- 
ting with the intercalar. A large flange unites the two limbs posteriorly and forms a 
medial facet for the articulation of the supracleithrum. The lateral line canal enters 
the post-temporal through a foramen in the anterior region of the supracleithral 
facet and the canal passes through the bone to open antero-laterally. 

The supracleithrum is large, flattened and broad with a thickened anterior edge. 
The lateral line canal enters the bone mid-way down the posterior edge. The blunt 
ventral end of the supracleithrum articulates with the lateral face of the cleithrum. 

The cleithrum is unusual in that it has no noticeable dorsal extent above the 
attachment of the endoskeletal girdle. (This character is also noticed in other 
myctophiforms, e.g. Chlorophthalmus.) The posterior part of the cleithrum forms a 
small plate-like expansion ornamented with a V-shaped groove on its lateral face. 
The dorsal postcleithrum attaches to the internal face of this posterior extension. 
Ventrally the cleithrum extends anteriorly and is unexpanded but the anterior edge 
inclines medially. The endoskeletal girdle attaches to an internal flange from the 
cleithrum. 

The scapula is a small, roughly rectangular bone, thin for the most part, but with 
a thickened posterior edge. It joins the internal cleithral flange dorsally and the 
coracoid ventrally. A large oval foramen is present in the central region of the 
scapula. On the thickened posterior edge of the scapula there is a smooth saddle- 
shaped facet for the articulation of the anteriormost fin ray. 

Ventrally the coracoid is produced into a long slender process which contacts the 
anteriormost tip of the cleithrum. A large fenestra is left between the coracoid and 
the cleithrum. The postero-ventral tip of the coracoid projects back as a short 
pointed process below the fin insertion. 

Two postcleithra are present, the larger dorsal one has a thickened anterior edge 
but is thinner and rounded posteriorly. The ventral postcleithrum is a small slip 
of bone attached to the antero-ventral edge of the dorsal postcleithrum. 

The pectoral fin consists of 13 rays, all of which are segmented distally. The first 
ray is the longest and both it and the second ray are unbranched, but the remainder 
of the rays are branched. The rays are supported by four radials which contact 
the thickened posterior edge of the scapula and the scapulo-coracoid suture. The 
radials are hourglass shaped and increase in size ventrally. 

Pelvic girdle and fin. The pelvic girdle is made up of two pelvic bones connected 
in the mid-ventral line. Each bone is large with a considerable horizontal extent, 
tapering anteriorly to a point. The fin is inserted on a thickened postero-lateral 
condyle which is merely a continuation of the posterior edge of the bone. The 
condyle is supported by a lateral strengthening ridge. The medial part of the 
posterior edge of the bone, internal to the insertion of the last pelvic ray, is drawn 
out posteriorly into a very long, narrow, blunt posterior process. Weitzman (1967 : 



ESPECIALLY MYCTOPHOIDS 



217 






to 

s 



6 



218 UPPER CRETACEOUS TELEOSTS 

513, fig. 6) shows the pelvic girdle of Salmo and illustrates small posterior cartila- 
ginous processes from the ' ischial plate '. These processes obviously correspond 
to the large ossified processes of Aulopus. 

The pelvic fin is composed of 9 rays all of which are segmented distally. The first 
ray is unbranched, the next three are simply bifurcated whereas the remaining five 
rays are normally branched. Gosline, Marshall and Mead (1966 : 5, fig. 3) have 
noticed the fusion of the ventral half of the innermost pelvic ray with the inner 
radial element. Like the structure of the palatine bone, however, this feature is of 
variable occurence in both salmoniforms and myctophiforms. Also it is a feature 
which is not easily seen in fossil forms and as such has been left out of consideration 
here. 

Vertebral column. The vertebral column is shown in the restoration, Text-figure 
92. It consists of 50 vertebrae, of which 14 are caudal. The centra are small, 
slightly longer than deep and mesially constricted. Each precaudal centrum bears 
a pair of small ventro-lateral transverse processes. These gradually increase in 
length towards the posterior end of the precaudal region. The transverse processes 
support thin, needle-like pleural ribs which are longest anteriorly but decrease in size 
posteriorly. The first precaudal vertebra articulating with the occipital condyle 
does not possess either transverse processes or pleural ribs. The neural arches along 
the entire length of the column are well ossified and intimately connected with the 
centra. The neural spines arise towards the posterior end of each arch. The 
neural spine of the first vertebra has become fused with the first accessory radial to 
produce an expanded plate of bone extending upwards in the mid-dorsal line. This 
plate extends back and tapers to end above the fifth precaudal vertebra. Two or 
three additional, unexpanded accessory radials occur posterior to the first and are 
ligamentously attached to the neural spines of precaudal vertebrae 3, 4 and 5. 
Prezygapophyses and postzygapophyses occur throughout the column but are never 
particularly prominent. Epineurals and epipleurals are associated with the 
precaudal vertebrae. 

Median fins and tail. The median fins are shown in the restoration, Text-figure 
92. The dorsal fin is in the second quarter of the back and consists of 16 rays. 
None of the rays is excessively elongate, the longest being the second and third. 
The first ray is the shortest and is unbranched, the longer second one is also un- 
branched. All of the rays are segmented. The dorsal fin radials are long and thin 
except for the anteriormost one which is expanded into an anterior keel-like 
structure. 

The anal fin occurs towards the posterior end of the body, well behind the dorsal 
fin. Eleven rays are present and again none is excessively elongate. The third and 
fourth rays are the longest. The first three rays are unbranched. None of the 
radials is expanded. 

An adipose fin without any skeletal components is positioned above the centre of 
the anal fin. 

The caudal skeleton (Text-fig. 93) is composed of five vertebrae : two free preural 
vertebrae (preurals 2 and 3) ; one fused vertebra (preural 1 and ural 1) ; and 



ESPECIALLY MYCTOPHOIDS 219 

a terminal half-centrum (ural vertebra 2). Preural vertebrae 2 and 3 both bear 
haemal spines which are slightly expanded and support the hypaxial accessory 
caudal rays. Only preural vertebra 3 bears a neural spine, preural vertebra 2 
merely supports an expanded neural arch. The fused vertebra (preural 1 and ural 
1) bears the parhypural (haemal spine of preural one) and the first two hypural 
bones ventrally. The centrum itself is slightly upturned. The parhypural is 
expanded and only has a small area of attachment with the centrum and anteriorly 
overlaps the haemal spine of preural vertebra 2. Hypurals 1 and 2 are sutured 
together, with hypural 1 being the larger. The terminal half-centrum bears a series 
of four hypurals supporting the upper fin rays. The stegural is borne on the dorsal 




ph 
Fig. 93. Aulopus filamentosus. Caudal fin skeleton in lateral view. 

surface of the fused centrum and consists of an elongate postero-dorsally directed 
strut, with an elongated expansion anteriorly. A second and possibly a third 
uroneural, both rod-like, occur postero-dorsally immediately behind the stegural. 
Three epurals are associated with the dorsal edge of the stegural. 

Nineteen principal caudal rays are present of which 17 are branched. Accessory 
rays are present epaxially and hypaxially, there being 16 in all. In front of the 
accessory rays a single large flattened fulcral scale is present above and below the 
peduncle. 

Scales. The entire body is covered with a uniform coat of regularly shaped, 
feebly ctenoid scales. The lateral line scales are slightly enlarged and bear a median 
ridge. In the head region the scales extend on to the cheek and the posterior end of 
the mandible. Scales also cover the rear end of the skull-roof as far as the parietals, 
and the anterior and posterior edges of the preoperculum, operculum and sub- 
operculum. 



UPPER CRETACEOUS TELEOSTS 



The genus Aulopus described in the preceding pages represents the most primitive 
living representative of the Myctophiformes and many comparable features can be 
seen in the Cretaceous Sardinioides. The similarities can conveniently be listed : 

i. Neurocranium shallow with a roofed post-temporal fossa. 

2. Parietals in contact along the mid-dorsal line. 

3. Both orbitosphenoid and basisphenoid present. 

4. Antorbital present. 

5. Two supramaxillae present, although somewhat reduced in extent. 

6. Maxilla long and narrow except posteriorly where it is deepened. 

7. Premaxilla with ascending and articular processes and a very long 

alveolar arm. 

8. Supratemporal present above the opening to the post-temporal fossa. 

9. Pectoral fin on the flank, the girdle having lost the mesocoracoid arch. 

10. Pelvic fin composed of 9 rays. 

11. In the caudal skeleton preural vertebra one is fused with ural vertebra 

one. 

12. Ural vertebra two present as a small half-centrum. 

13. An adipose fin present (seen in Sardinioides monaster i from Sendenhorst, 

Westphalia). 

14. Basal fulcral scales present both above and below the caudal peduncle. 

15. Fin spines absent. 

16. Three or four accessory radials (interneurals) between the occiput and 

the dorsal fin. 

17. Epineurals present on all of the precaudal vertebrae and epipleurals 

present in abundance. 

18. Body scaling extends on to the opercular bones and certain of the head 

bones. 

These numerous similarities in overall construction represent the basic characters 
of the myctophiforms, the differences being few in number and of only minor signi- 
ficance. Firstly, Sardinioides has a narrow parasphenoid below the orbit, but in 
Aulopus the parasphenoid is greatly expanded. The supraorbital bone would 
appear to be absent in Sardinioides but is large and forms part of the orbital border 
in Aulopus. However it may be that what is termed an antorbital in the description 
of Sardinioides minimus (see p. 154) is a supraorbital. The ascending process 
of the premaxilla differs somewhat ; in Sardinioides (Text-fig. 70) it is long, but 
in Aulopus (Text-fig. 90) it is considerably reduced and rounded. In Sardini- 
oides the suboperculum is little expanded and the operculum is large, whereas in 
Aulopus the suboperculum has greatly expanded dorsally to form the whole posterior 
edge of the opercular cover, the operculum consequently having become greatly 
reduced. The pelvic fins, although possessing the same number of rays in both 
genera, are abdominal in position in Sardinioides (Text-fig. 71) but seem to have 
migrated to a sub-thoracic position in Aulopus (Text-fig. 92). Sardinioides has a 
pelvic splint bone (Gosline, 1961 : 18) on the outer surface of the dorsal half of the 



ESPECIALLY MYCTOPHOIDS 221 

first pelvic ray, but this structure is absent in Atdopus. The branchiostegal ray 
number is variable. There are 8 or 9 rays in Sardinioides, but Aulopus has approxi- 
mately 15. The vertebral column is a variable feature in that the elements are 
incompletely fused in Sardinioides but are much more closely united in Aulopus. 
Although the total vertebral count is different (30 in Sardinioides, 50 in Aulopus), 
the number of caudal vertebrae is very similar, with 16 in Sardinioides and 14 in 
Aulopus. Both genera have feebly ctenoid scales, but in Sardinioides cycloid scales 
are also present. Although the basic composition of the caudal skeleton is the same 
in the two genera, that of Sardinioides (Text-fig. 72) has a fairly wide gap between 
the second and third hypurals, whereas in Aulopus (Text-fig. 93) hypurals two and 
three contact each other. 

These differences are only minor ones and the two genera are extremely closely 
related although being considerably separated in time. It is difficult to derive 
Aulopus directly from Sardinioides because of the possible absence of a supraorbital, 
the small number of branchiostegal rays and the presence of a gap between the 
second and third hypurals in Sardinioides. The hypural gap is also seen in the 
Ctenothrissiformes (Text-fig. 77), in certain polymixioid beryciforms (Patterson, 
1968b), and in the Paracanthopterygii (sensu Greenwood, et al., 1966). No obvious 
functional significance can be attached to this feature, the fin outline remaining 
the same irrespective of whether the gap is present or absent. Many of the more 
advanced myctophiforms exhibit a slight gap but never to the extent shown in 
Sardinioides. If this is in fact a variable feature, then it would not on its own 
exclude Sardinioides from the ancestry of Aulopus. 

Thus it would appear that Sardinioides although not directly ancestral to Aulopus 
is extremely close to this ancestry. Both genera represent equivalent basal forms 
from which most, if not all, of the present day myctophiforms can be derived. 

The two genera Acrognathus and Nematonotus which were included in Woodward's 
(1901) family Scopelidae both show strong affinities with the myctophiforms. The 
skull-roof of an undetermined species of Nematonotus (B.M.N.H. specimen P. 48681) 
is shown in Text-figure 94, and it is very similar to those of both Sardinioides and 
Aulopus (Text-fig. 86). Apart from the skull-roof there are certain minor differences. 
The head and body of Nematonotus (Text-fig. 95, Nematonotus bottae) are somewhat 
deeper than in Sardinioides, and the premaxilla has a very long ascending process 
and a large articular process. One major difference occurs and this concerns the 
maxilla which in Nematonotus enters the gape behind the premaxilla, a feature 
not seen in any myctophiforms, and the maxilla is toothed. Corresponding to the 
entry of the maxilla into the gape, the bone itself is deeper and supports two supra- 
maxillae which are considerably larger than those in Sardinioides. Like Aulopus, 
but unlike Sardinioides, there is no gap between the second and third hypurals. 
The base of the first uroneural is forked in Nematonotus bottae (Text-fig. 96), this 
being a primitive feature seen in such forms as Elops (Text-fig. 80) and to a very 
slight extent in Salmo (Text-fig. 81), but in neither Sardinioides nor Aulopus (see 
also Patterson, 1968a : 226). 



UPPER CRETACEOUS TELEOSTS 



The major feature in which Nematonotus differs from the myctophiforms is the 
toothed maxilla. Its upper jaw is intermediate in form between the myctophi- 
form Sardinioides and the ctenothrissiform Pattersonichthys. On this basis the genus 
Nematonotus cannot be satisfactorily included within the Myctophiformes and 



mes 




inf s.c 



o.inf. s c 



Fig. 94. Nematonotus sp. Neurocranium in dorsal view. The broken line on the right 
side of the figure indicates the course of the sensory canals. From B.M.N.H. number 
P. 48681. Middle Cenomanian, Hajula, Lebanon. 



ESPECIALLY MYCTOPHOIDS 



223 




CD 



0) 






ID 



!> 

Q 









-- 1 



UPPER CRETACEOUS TELEOSTS 



because of its abdominal pelvic fins cannot be included within the Ctenothrissi- 
formes. Thus Nematonotus emphasizes the basal convergence of the ctenothrissi- 
form and myctophiform lineages, and for the time being is best included in the 
Salmoniformes as a separate suborder, Nematonotoidei. 



ep3 ep2 



n.a.pu2 




pul+ul 



5mm 



Fig. 96. Nematonotus bottae (Davis). Caudal fin skeleton in lateral view. 



A further genus considered by Woodward (1901) to be a myctophiform is Acro- 
gnathus. In many respects this resembles Sardinioides, notably in respect of the 
caudal skeleton, but the maxilla, although apparently untoothed, is expanded 
posteriorly and enters the gape behind the premaxilla. The maxilla also supports 
two large supramaxillae. The large size of the basal fulcral scales recalls Sardinioides 
and Aulopus, but not the Ctenothrissiformes in which the fulcral scales are reduced. 
An adipose fin could not be made out in Acrognathas but the pelvic fins are abdominal. 
Perhaps Acrognathus is best included with Nematonokis in the suborder Nematono- 
toidei, which lies very close to the group from which both the Myctophiformes and 
the Ctenothrissiformes were derived. 

The last of the Cretaceous genera considered by Woodward (1901) to be a scopeloid 
is Sardinius. The species Sardinius cordieri von der Marck (Text-fig. 97) certainly 
shows several myctophiform characters which are comparable with Sardinioides and 
Aulopus. For instance the maxilla is excluded from the gape, the ascending process 
is abbreviated (as in Aulopus), the suboperculum is large and the operculum is 
small, preural vertebra one is fused with ural vertebra one in the caudal skeleton, 
nnd the neurocranium is shallow. 



ESPECIALLY MYCTOPHOIDS 



225 













& 



o 
u 
aj 



^S 



s 
s 






226 UPPER CRETACEOUS TELEOSTS 

Certain features shown by Sardinius would appear to be in a less advanced condi- 
tion than those shown in the Sardinioides-Aulopus complex, and yet others appear 
to be more advanced. The more primitive features are : 

i. Pectoral fin low on flank with as many as 22 rays. 

2. Pelvic fins abdominal with 11 rays. 

3. Five or six accessory radials present in advance of the dorsal fin. 

The more advanced features are : 

1. No posterior expansion of the maxilla. 

2. No supramaxillae. 

3. Much widened gape with the jaw articulation positioned posterior to the 

occiput. 

4. No fulcral scales at the base of the tail. 

5. No pelvic splint bones. 

6. Markedly ctenoid scales. 

Sardinius from the above considerations seems to have been possibly an early 
offshoot from either the myctophiform lineage or the group of Salmoniformes from 
which the Myctophiformes were derived. It is obviously much closer to the mycto- 
phiforms than to any of the other salmoniform groups so far considered, e.g. encho- 
dontoids, halecoids. The evolution of Sardinius appears to have paralleled that of 
the myctophiforms in that some basic myctophiform characters occur together with 
some more advanced characters while several ' premyctophiform ' features are still 
retained, the overall impression being not unlike the Myctophidae. 

An additional genus appended here is Volcichthys, with the type and only species 
Volcichthys dainellii d'Erasmo (1946), from the Lower Cenomanian of Volci in Comen. 
This genus was placed by d'Erasmo (1946) in the Enchodontidae, but it most 
certainly has no affinity with these. From a consideration of Text-figure 98, it can 
be seen to have a certain degree of similarity with Sardinius. Both genera have a 
long narrow dermal upper jaw, and supramaxillae are absent, but Volcichthys has 
the maxilla entering the gape and bearing teeth. Possibly the two might be related 
and thus Volcichthys can be considered to have a similar origin to that already 
proposed for Sardinius. 

The basal stock of the myctophiforms, as represented by Sardinioides, was present 
at the beginning of the Upper Cretaceous and it has been shown earlier (p. 203, 
table 1) that great similarities exist between this stock and the ctenothrissiforms. 
Patterson (1967a) has shown that the ctenothrissiforms gave rise to the beryciforms 
in the Albian or thereabouts. Thus at the opening of the Upper Cretaceous both the 
myctophiforms and the beryciforms were present and increasing, whilst the cteno- 
thrissiforms were rapidly declining. Due to the incompleteness of the fossil record 
little indication can be gained concerning the time of origin of many of the recent 
myctophiform groups (Patterson, 1967b). Aulopus has already been considered in 
some detail and is little changed from its Cenomanian ancestors. A second recent 



ESPECIALLY MYCTOPHOIDS 



227 



■^ 



228 UPPER CRETACEOUS TELEOSTS 

genus, Chlorophthalmus, although somewhat more advanced than Aulopus is still little 
different from Sardinioides. The remainder of the recent forms are much more 
specialized, particularly for life in deeper waters. Harry (1953 : 244) has identified 
paralepidids in the Miocene of California. These are somewhat more advanced than 
Chlorophthalmus, and are on the line which culminates in the present day Anotopterus 
and Lestidium. Arambourg (1925) has described a Miocene myctophoid fauna from 
Licata in Sicily, and his specimens are closely related to the present day Myctophidae. 
The Californian Miocene deposits certainly seem to be deep water (Crane, 1966 : 20), 
as do the Sicilian ones. 

The colonization of a deep sea environment has been considered by Marshall 
(1963 : 189) who stated that radiation into the central water masses would have been 
from a relatively rich to a poor food producing area. The cause of this migration 
into deeper water can be accounted for if the factor of competition is considered. 
During the Upper Cretaceous the beryciforms and subsequently the perciforms 
were radiating and occupying many niches already occupied by salmoniforms 
and myctophiforms. The acanthopterygians with their more versatile jaw structure, 
their increased manouevrability and the development of protective mechanisms 
as well as offensive mechanisms on both fins and scales seem to have provided 
overwhelming competition for many groups. The Myctophiformes, however, 
seem to have had sufficient evolutionary plasticity to become adapted to a new 
environment, that of deeper water. Thus the myctophiforms were able to develop 
phosphorescent organs, increase the size of the gape, lose the air bladder and 
decrease the amount of ossification and squamation. Together with these particular 
specializations the myctophiforms show other trends which are present in the fossil 
salmoniforms, e.g. enchodontoids and halecoids, and which mirror features seen 
throughout the Acanthopterygii. These include the loss of the orbitosphenoid, 
basisphenoid, antorbital and supraorbital ; the separation of the parietals ; the 
loss of supramaxillae, and finally the loss of the roof to the post-temporal fossa. In 
the myctophiforms this reduction is obviously useful in that it reduces the density of 
the body, but this effect would appear to be only a secondary one. Throughout the 
Protacanthopterygii and Acanthopterygii, irrespective of habitat and mode of life, 
the more advanced and specialized members undergo these reductions. 

In a discussion of the Myctophiformes some consideration must be given to the 
Paracanthopterygii of Greenwood, el al. (1966). This group represents a parallel 
spiny-finned radiation to the Acanthopterygii and contains the percopsiforms, 
batrachoidids, lophioids, ceratioids and gadoids. Greenwood, el al. (1966 : 371) 
proposed that this group might have been derived from some Cretaceous myctophi- 
form such as Sardinioides. They tentatively proposed that a ' paraberycoid ' 
radiation might have occurred, this hypothetical radiation being distinct from the 
berycoids and the two being independently derived from the Salmoniformes. It is 
quite possible that the paracanthopterygians represent forms which arose from the 
myctophiform lineage at about the time that the beryciforms arose. Some degree 
of substantiation for the alignment of the paracanthopterygians with the myctophi- 
forms is the presence of an adipose fin in both groups. The adipose fin is not 



ESPECIALLY MYCTOPHOIDS 229 

represented in the ctenothrissiforms or the beryciforms. A list of certain basic 
differences between the two lineages is given below : 

Myctophiformes Ctenothrissiformes 

Paracanthopterygii Beryciformes 

1. Very long alveolar process of the premaxilla Alveolar arm much shorter 

under the maxilla. 

2. No teeth on the maxilla. Teeth on maxilla in ctenothrissids and 

certain berycoids. 

3. Supramaxillae small or absent. Supramaxillae always present and large. 

4. No subocular shelf on infraorbital bones. Subocular shelf present on infraorbital 

bones. 

5. Adipose fin present in myctophiforms Adipose fin never present. 

and Percopsidae. 

6. Undeepened body. Body much deepened in the beryciforms. 

Together with these differences, Greenwood, et al. (1966, p. 388) have pointed out 
several differences concerned with the levator maxillae muscles and the recurrent 
facial nerve. 

The condition of the caudal skeleton in the paracanthopterygians is unusual 
(Gosline, 1961, 1963 ; Rosen, 1962). Instead of the normal fusion of the ural 
vertebrae with the first preural vertebra followed by subsequent fusion of the upper 
hypurals to this compound centrum, as has occurred in most higher teleosts, the 
paracanthopterygians have the upper hypurals fused to the second ural vertebra 
and this vertebra retains its intervertebral articulation with the fused first preural 
and ural vertebrae. This condition is approached by certain of the advanced 
myctophiforms such as Neoscopelus, but by none of the beryciforms. Sardinioides 
is also similar and additionally has a gap between the second and third hypurals, a 
feature noticed in many paracanthopterygians. Among other fossil forms Volci- 
chthys, noticed earlier (p. 226, Text-fig. 98), has a caudal skeleton closely approaching 
that of certain amblyopsids. The second ural vertebra has fused with the upper 
hypurals, while the fused preural and ural vertebrae support the fused first two 
hypurals as well as the parhypural. A prominent gap is also present between the 
second and third hypurals. Possibly Volcichthys and the related Sardinius are 
close to the ' paraberycoid ' hypothetical assemblage, additional support being 
added by the presence of well developed ctenoid scales in Sardinius. 



Order CTENOTHRISSIFORMES 

The new genus Patter sonichthys has been erected to contain several small specimens 
from the Middle Cenomanian of Hajula, Lebanon. The genus clearly belongs in the 
order Ctenothrissiformes showing those characters enumerated by Patterson (1964 : 
244) for the order. The primitive features visible in Patter sonichthys are as follows : 

1. Antorbital present. 

2. Orbitosphenoid present. 

3. Premaxilla small, but with a developing ascending process. 



230 



UPPER CRETACEOUS TELEOSTS 



fin rays are, however, 
initially indicated that 



4. Large maxilla supporting two large supramaxillae 

5. Incompletely fused vertebral elements. 

6. Numerous epineurals. 

7. Parietals meet in the mid-line of the skull-roof. 

8. Roofed post-temporal fossa. 

9. 17 branched caudal fin rays. 
10. No true fin spines. (The anteriormost dorsal 

unsegmented and spinous.) Woodward (1901 
Pateroperca libanica had three slender spines at the anterior end of the 
dorsal fin. Patterson (1964 : 371) was of the opinion that these rays 
were segmented distally and not spinous. In Pattersonichthys the first 
two dorsal rays although composed of ray halves basally show no 
evidence of having been segmented. 

When compared with the genus Ctenothrissa, the genus Pattersonichthys shows 
several characters which appear to be more primitive : 

1. Parietals meet for their entire length along the mid-dorsal line, and the 

supraoccipital is small. 

2. Post-temporal fossa roofed. 

3. Head bones unornamented. 

4. Short ascending process on the premaxilla. 

5. Maxillary head simple in form. 

6. Ceratohyal only slightly deepened. 
Pelvics sub-thoracic. 
Scales mainly cycloid. 
Body undeepened. 

Thus the genus Pattersonichthys is more generalized than Ctenothrissa and lies 
closer to Aulolepis and Pateroperca. Aulolepis, Pateroperca and Pattersonichthys are 
all nearer to the direct ancestry of the Acanthopterygii (sensu Greenwood, et al., 
1966) than is Ctenothrissa itself (see Patterson, 1964). The Ctenothrissiformes as a 
group present a generalized picture with few specializations (except in Ctenothrissa) . 
These more generalized ctenothrissiforms can clearly be derived from the basal 
salmoniform stock, Pattersonichthys differing from this stock in a few details. These 
include the absence of a supraorbital ; reduction in the number of branchiostegal 
rays (9) ; reduction in vertebral number ; pelvics sub-thoracic ; mesocoracoid 
arch absent ; and only one free ural vertebra. These points have already been listed 
in Table 1 during the consideration of the Myctophiformes. 

The first three of these points are relatively minor changes, whereas the latter three 
features represent more fundamental changes in organization, correlated with a 
greater efficiency in movement. The loss of the mesocoracoid arch is due to the 
migration up the flank of the pectoral fin, from which position the pectoral fin may 
act as a brake. The forward migration of the pelvic fins is correlated with the use of 
the pectorals as brakes, the pelvics being used as stabilizers (Harris, 1938 : 37 ; 
Patterson, 1964 : 452). The consolidation of the caudal skeleton is the first of 



ESPECIALLY MYCTOPHOIDS 231 

many steps to reduce the number of moveable parts in the caudal fin. The tail is 
only required to move in one plane and it would seem that fusion of component 
parts would produce a more effective fin. In the most advanced perciforms the 
consolidation has reached an end point in which the caudal skeleton is only formed 
from one or two large plates. 

An historical review of the Ctenothrissiformes has been given by Patterson 
(1964 : 241). In this work Patterson considered three genera, Ctenothrissa, 
Aulolepis and Pateroperca, which he put into two families : 

Family Ctenothrissidae. Ctenothrissa. 

Family Aulolepidae. Aulolepis and Pateroperca. 

In his discussion (Patterson, 1964) of the affinities of Ctenothrissa he concluded 
that " Ctenothrissa is a form which has evolved towards the acanthopterygians in 
a number of characters (thoracic pelvics, no mesocoracoid, ctenoid scales, and deep 
body in particular), while still showing unmistakeable signs of an elopoid ancestry 
(small supraoccipital, foramen for orbital artery, antorbital, toothed maxilla, and 
incompletely fused vertebral elements), yet has become specialized in ways which 
exclude it from the direct ancestry of the acanthopterygians ". However, Aulolepis 
shows several differences to Ctenothrissa, but as Patterson (1964) points out, these 
are only in degree. Pattersonichthys, like Aulolepis and Pateroperca, is more general- 
ized than Ctenothrissa and is included in the family Aulolepidae. 

One salient feature of Pattersonichthys not found in any other ctenothrissiform 
genus is the roofed post-temporal fossa. The absence of a roofed fossa excludes both 
Aulolepis and Pateroperca from the direct ancestry of the Beryciformes, since a 
partially roofed post-temporal fossa is still present in the early beryciforms (Hoplop- 
teryx, Trachichthyoides , Patterson, 1964 ; Lissoberyx, Caproberyx, Patterson, 1967a). 
The presence of a roofed fossa in Pattersonichthys suggests that it is nearer to the 
direct line of ancestry of the Beryciformes than is either Ctenothrissa, Aulolepis or 
Pateroperca. 

Patterson (1967a : 106) has suggested that the ancestral beryciform stock would 
have possessed the following characteristics : 

1. Of small size with unornamented head bones. 

2. High supraoccipital crest from a moderately large supratemporal fossa. 

3. Reduced antorbital. 

4. Toothed maxilla. 

5. Scaly operculum. 

6. Partial roof to the post-temporal fossa. 

These characters are all present in the Ctenothrissiformes but in mosaic occurrence 
Pattersonichthys for example shows 1, 3, 4 and 6 but not 2 and 5. 

The basic caudal pattern shown by the Ctenothrissiformes (Text-fig. yy) and 
Beryciformes (Patterson, 1968b) is similar to that already seen in the Myctophiformes 
(Text-fig. 72) . This type of caudal skeleton was also present in several of the fossil 
salmoniform groups mentioned in this work, the Enchodontoidei (Text-fig. 42), 
Halecoidei (Text-fig. 64) and Cimolichthyoidei. 



232 UPPER CRETACEOUS TELEOSTS 

All of the groups mentioned above except the Beryciformes are included within the 
Protacanthopterygii (sensu Greenwood, et al., 1966), whose basal stock is represented 
by the salmonoids. From a caudal skeleton like that of Salmo (Text-fig. 81) all of 
these other groups can be derived by the fusion of preural vertebra one with ural 
vertebra one. The basic caudal composition of the Enchodontoidei, Cimolich- 
thyoidei, Halecoidei, Myctophiformes, Ctenothrissiformes and Beryciformes is as 
follows : 

1. Preural vertebra one fused with ural vertebra one. 

2. Ural two present as a terminal half-centrum. 

3. Six hypurals. 

4. Three epurals. 

5. A stegural formed from uroneural one and neural arch components. 

6. A second uroneural behind the stegural. 

A variable feature in the caudal skeleton of these six groups is concerned with the 
neural arch and spine of the second preural vertebra (i.e. the first free preural 
vertebra). In the Beryciformes, Patterson (1968b) states that the polymixoids 
and some dinopterygoids retain the neural spine of preural vertebra two while in 
the remaining dinopterygoids and the berycoids the spine is lost. In the Cteno- 
thrissiformes, Ctenothrissa, Aulolepis and Patter sonichthys have lost the neural spine, 
and it is only retained in Pateroperca. 

Thus in respect of the post-temporal fossa Patter sonichthys is closest to the ancestry 
of the beryciforms, but in caudal structure is further removed than Pateroperca. 
Patterson's (1964) conclusions are still valid in that no so far described cteno- 
thrissiform could have been directly ancestral to the beryciforms, although this 
ancestry clearly lies within the ctenothrissiform grade. 

Patterson (1964 : 465) has also used the distribution in time of the ctenothrissi- 
forms as supporting evidence that the group is near the ancestry of the Acantho- 
pterygii as a whole. The more generalized members of the Ctenothrissiformes, 
which are rare in the Cenomanian, have died out by the Turonian, being replaced by 
the more specialized beryciforms which they had given rise to in the Albian or earlier 
(Patterson, 1967a : 107). The more specialized Ctenothrissa, which parallels the 
beryciforms to a certain extent, seemed able to compete with the berycoids and 
continued into the Turonian. This distribution is to be expected if the more 
generalized ctenothrissiform stock (represented by Aulolepis, Pateroperca and 
Patter sonichthys) had given rise to the Beryciformes. 

Patterson (1964) has evaluated the merits of deriving the Acanthopterygii from 
the Ctenothrissiformes or from the Myctophiformes, and has shown conclusively that 
the ancestral stock lies closer to the ctenothrissiforms. He further indicated that 
the Ctenothrissiformes were not closely related to the generalized Myctophiformes 
although both groups had reached the same stage of evolution. As has already been 
indicated in the discussion of the myctophiforms both of these groups are far more 
closely related than Patterson (1964) supposed. Furthermore it has also been 
indicated that the bases of both groups are little removed from a basal salmoniform 



ESPECIALLY MYCTOPHOIDS 233 

stock. For example Patter sonichthys delicatus differs from Sardinioides minimus in 
the following features : 

1. A small broad ascending process with a short alveolar arm on the pre- 

maxilla. 

2. Long, stout maxilla which is extensively toothed, forming two-thirds of 

the gape. 

3. Two very large supramaxillae. 

4. Unexpanded suboperculum. 

5. Sub-thoracic pelvic fins. 

6. Cycloid scales which do not extend on to any of the skull bones. 

7. Fulcral scales reduced in extent. 

Apart from these points the two genera are practically identical in all respects. 
Thus two groups (Ctenothrissiformes and Myctophiformes) which are ordinally 
separated have basal members which converge strongly, although the ctenothrissi- 
form and myctophiform lineages are evident in the convergent stocks. Patterson 
(1967a : 104) has considered a similar basal convergence of the trachichthyid and 
holocentrid berycoids at the base of the Cenomanian. 

Although it has now been indicated that the basal myctophiforms and cteno- 
thrissiforms are on equivalent levels of structural organization, the evolutionary 
potential within either group would appear to be different. The ctenothrissiforms 
were endowed with the ability to produce a higher level of organization, the Acantho- 
pterygii (sensu Greenwood, et al., 1966). The evolutionary potential in the mycto- 
phiform lineage would appear to have been limited solely to producing diversified 
members of the same structural level of organization, although certain of the 
descendants are extremely specialized for particular habitats. This viewpoint, 
however, excludes the possibility of the basal myctophiforms having given rise to the 
Paracanthopterygii (the parallel spiny-finned radiation) as Greenwood, et. al., 
(1966) supposed. If the myctophiforms did give rise to the Paracanthopterygii then 
the evolutionary potential of the myctophiforms and ctenothrissiforms was similar. 



IV. TELEOST EVOLUTION WITHIN THE CRETACEOUS 

Before the evolution of the teleosts in the Cretaceous can be considered it is 
necessary to give some discussion of the evolution of the teleosts as a whole. The 
first question to be answered is whether or not the teleosts represent a monophyletic 
group or a polyphyletic one. Patterson (1967c : 95) has given this question some 
thought but indicates that one must first settle upon the definition to be attached 
to the term polyphyletic. Recently most authors have considered that the teleosts 
had their origin within the halecostomes. Thus monophyly occurs at the level 
of the division Halecostomi. If, however, within the halecostomes it is found that 
more than one family had given rise to the teleosts, then polyphyly is present at 
the familial level. This point is discussed in some detail by Patterson (1967c), 
who is of the opinion that no grounds have so far been put forward which detract 



J34 UPPER CRETACEOUS TELEOSTS 

from the idea that the teleosts are monophyletic at the level of the family Pholido- 
phoridae. 

The division Halecostomi represents a transition grade between the basal holo- 
steans and the teleosts and was first used by Woodward (1932). The central stock 
of the halecostomes would appear to be the family Pholidophoridae, and Gardiner 
(i960) has convincingly shown that this family can be derived directly from the early 
Parasemionotiformes. The family Pholidophoridae is normally grouped with five 
other families in the order Pholidophoriformes (see Griffith and Patterson, 1963, for 
consideration of these families). The Pholidophoridae appears to be closest in 
structural organization to the teleosts whereas the other families are more distantly 
removed. 

The Pholidophoridae sensu stricto have recently been extensively examined by 
Nybelin (1966) who agrees with Gardiner (i960) that they were derived from a 
parasemionotid-like ancestor. In confirmation Griffith and Patterson (1963 : 37) 
stated that the parasemionotids are almost ideal intermediates between the chon- 
drosteans and the halecostomes. 

The division Halecostomi also includes the order Leptolepiformes, and this order, 
together with the Pholidophoriformes, presents a collection of families which 
exhibit a mosaic of halecostome and teleostean characters (mosaic evolution is used 
in the sense that de Beer, 1954, and Schaeffer, 1965, indicated, viz. lineages of com- 
mon ancestry exhibiting various combinations of primitive, intermediate and ad- 
vanced characters). Schaeffer (1965 : 322), in reference to the origin of the teleosts, 
indicated that every organism is a mosaic whether or not it is associated with a 
transition between grades or with an adaptive radiation within a grade. The latter 
point has been amply illustrated in the foregoing discussions where radiations 
within the Salmoniformes have been considered. It is hoped that the former point 
will become apparent throughout this section. 

This mosaic nature of the halecostomes makes it difficult to define a ' teleost '. 
Patterson (1967c) has argued the ' pros and cons ' of the dilemma and whether or not 
it is advisable to recognize the Teleostei as a taxonomic group. Lehman (1966) and 
Berg (1940) for example do not recognize a division Teleostei, but Patterson 
(1967c : 95) finds that a category Teleostei is necessary and that the base of this 
group is within the leptolepids. 

The teleost characters shown by the halecostomes are related to changes in 
respiratory, feeding and locomotor mechanisms and include : 

1. Loss of enameloid tissue (ganoine) from scales and dermal bones. 

2. Thinning of scales and loss of the peg and socket joint. 

3. Ossification of the vertebrae as single units. 

4. Development of numerous intermuscular bones. 

5. Compact caudal skeleton with two hypurals to the lower caudal lobe 

supported on ural vertebra one. 
These features all indicate the increased flexibility of the body and compensate in 
part for the increased mechanical stresses built up by the musculature (Nursall, 1956). 
One primitive character of the halecostomes is the presence of enamel as a thin layer 



ESPECIALLY MYCTOPHOIDS 235 

covering the scales and dermal bones in pholidophorids (Nybelin, 1966). The loss 
of enamel has made possible the sinking inwards of dermal bones of the skull and the 
production of uroneurals in the caudal fin skeleton in teleosts, This latter feature, 
concerning the internal uroneurals attached to the ural vertebrae, has been used by 
Patterson (1967c) provisionally as a diagnostic feature of teleosts. 

The leptolepids have convincingly been shown to have their origin within the 
Pholidophoridae (Rayner, 1948 : 338), and several authors (Rayner, 1948 : 340 ; 
Griffith and Patterson, 1963 : 40) have proposed that the Leptolepidae gave rise to 
the remainder of the teleosts. Other authors have supposed only the clupeoids to 
be direct derivatives of the leptolepids. Gardiner (i960 : 351) has indicated that 
both the Clupeidae and Chirocentridae can be derived from the leptolepids, all three 
groups having no bone cells in their scales, whereas elopoids and salmonoids, which 
still retain bone cells in the scales, must have evolved from pholidophoroids. 
Greenwood, et al. (1966) have stressed that the Clupeomorpha have several unique 
characters which have so far not been elucidated in the leptolepids. Patterson 
(1967c) has considered the origin of the Clupeomorpha in detail and has refuted 
Bardack's (1965) hypothesis concerning the derivation of the Chirocentridae from 
forms like the ichthyodectids and the Upper Jurassic Thrissops and Allothrissops. 
The origin of the Clupeidae according to Saint-Seine (1949) from a form like Lepto- 
lepis coryphaenoides has also been refuted by Patterson (1967c). Woodward 
(1942b : 908) postulated the derivation of Dipiomystus from a pholidophoroid and 
not from a leptolepid because of the presence of ridge scutes behind the pelvics in the 
diplomystids and in some pholidophorids. This character of Woodward's (1942b) is 
insignificant, but nevertheless his idea of deriving the clupeoids from a pholidophorid 
and not a leptolepid is sound. Arambourg (1954), in describing the new genus 
Clupavus, related it to the clupeoids, in particular the Dussumieriidae. Patterson 
(1967c) has convincingly shown this to be incorrect and has demonstrated the 
closeness of Clupavus to the leptolepids, and their dissociation from the clupeo- 
morphs. Greenwood, et al. (1966 : 360) are in agreement with Patterson (1967c), 
and both propose a connection between Clupavus and Leptolepis in a convergent 
evolutionary trend to the Clupeomorpha. This provides an instance of what 
Schaeffer (1965 : 322) maintained when he stated, " mosaics which in retrospect 
are involved in a transition to a higher level, may be modified in much the same way 
in a number of related lineages ". 

The Elopiformes represent the most primitive of living teleosts and this group 
retains the greatest number of halecostome characters. Gardiner (i960 : 351) has 
indicated that on scale structure alone the Elopiformes could be derived most 
satisfactorily from a pholidophoroid but not from a leptolepid. Saint-Seine (1949) 
has described a late Jurassic ' protelopid ' Eoprotehps, but Gardiner (1967b) con- 
siders this to be a halecostome, and Patterson (1967b) also indicates that it is not an 
elopoid. Bertin and Arambourg (1958 : 2200) had previously appended Eoprotehps 
to the family Leptolepidae. 

An elopid caudal skeleton has been described from the late Jurassic by Nybelin 
(1963) and it is probable that the elopiform lineage was in existence throughout 



236 UPPER CRETACEOUS TELEOSTS 

the early Jurassic, competing with the leptolepids and pholodophorids, as well as 
the majority of the Holostei. The elopids and the leptolepids thus appear to 
represent independent attainments of the teleost level, if the leptolepids are considered 
to be teleosts. 

The similarities between the Pholidophoridae, Elopidae and Leptolepidae can 
conveniently be listed : 

i. No articular or ascending processes on the premaxilla. 

2. Rostral elements with a bone enclosed ethmoidal commissure. 

3. Several supraorbitals. 

4. Fringing fulcral scales. 

5. Gular plate present. 

6. Many posterior branches from the preopercular sensory canal. 

7. No stegural. 

8. No adipose fin. 

The Salmoniformes represent a further teleost stock which shows a wide divergence 
from the Elopiformes. This divergence has been indicated by Greenwood, et at. 
(1966) who have given superordinal rank to these two phyletic lineages. The basal 
salmoniform stock exhibits many differences not only from the elopiforms, but from 
the pholidophorids and the leptolepids. These differences include : 

1. Articular process on the premaxilla. 

2. No separate dermal rostrals, no bone enclosed commissure. 

3. One small supraorbital. 

4. No fringing fulcra. 

5. No gular plate. 

6. No posterior branches of preopercular sensory canal. 

7. Stegural present. 

8. Adipose fin present. 

Several further dissimilarities are seen between the Recent forms Megahps and 
Salmo. Megalops has an otophysic connection, a leptocephalus larva, parasphenoid 
teeth and a subtemporal fossa. 

Discussion of the origin of the Salmoniformes is rendered difficult by the complete 
absence of fossil material below the Albian stage of the Lower Cretaceous. However, 
any one of these features in which the Salmoniformes differ from the Pholidophoridae 
could have been derived directly by reduction or loss through intermediate forms, as 
yet unknown, presumably within the Jurassic. For example, Nybelin (1967 : 244) 
has shown that the bone enclosed ethmoidal commissure of the elopoids is homologous 
with the canal system embedded in the soft tissue of the snout of clupeoids, and the 
pit-line on the snout of salmonoids. The commissure, then, in the salmonoids, has 
lost its bony surround together with any other rostral elements, except for the 
antorbital. Gardiner (i960 : 351) has said that the Salmonidae and Characinidae 
can be derived much more easily from a pholidophorid ancestor than from a lepto- 
lepid one, not just on scale structure, but on general disposition, in particular of the 
jaws. From my own observations I think it impossible to derive a basal salmoni- 
form from a leptolepid such as Leptolepis africana Gardiner (i960 : 314, fig. 48) 



ESPECIALLY MYCTOPHOIDS 237 

because of the enormous coronoid process of the mandible of Leptolepis, and the 
practically edentulous condition of the jaws as a whole. Other leptolepids however, 
e.g. Leptolepis coryphaenoides (Bronn) (Text-fig. 99D) and Leptolepis normandica 
Nybelin (Text-fig. 99C) have a lower coronoid and are provided with rather more 
teeth. The Pholidophoridae also exhibit a relatively large coronoid and small 
teeth (except Pholidophorus similis, Saint-Seine 1949, although Saint-Seine's 
reconstruction of this fish is to be viewed with some reservation). This does not 
imply that the salmoniforms were not derived from some pholidophorid, as the jaws 
are a particularly labile entity capable of easy modification. Thus it is conceivable 
that a leptolepid such as Leptolepis normandica may have provided the ancestral 
salmoniform stock. 

Possibly a trivial character, but one worth mentioning, is the suprapreopercular 
bone bearing the upper part of the preopercular sensory canal (formed from the 
fragmentation of the upper part of the preoperculum). This bone is seen in the 
salmonid Salvelinus (Norden, 1961), certain Gonorhynchiformes and many Ostari- 
ophysi (Greenwood, et al., 1966). Greenwood, et al. (1966) have indicated the 
extreme primitiveness of these three groups and place them near to the base of the 
protacanthopterygian radiation. Greenwood, et al. (1966 : 382) describe the 
ostariophysan suprapreoperculum as "an ossicle above the uppermost part of the 
preoperculum and partially surrounding the sensory canal ". It seems probable 
that since it is a canal bearing bone, the suprapreoperculum would have been present 
in the ancestral form. Fragmentation of the preoperculum occurs throughout the 
Chondrostei, including the Parasemionotidae (see Gardiner, 1967a : 200). In most 
cases it is the non-canal bearing anterior part of the preoperculum which fragments to 
produce the suborbital series. Less frequently the canal bearing portion of the 
preoperculum also fragments to give a smaller suprapreoperculum dorsally (Gardiner, 
1967a, fig. 2A). Such a condition is seen in Parasemionotus labor dei Piveteau (Text- 
fig. 99A) and the totally unrelated chondrostean Boreosomus gillioti Priem. No so 
far described pholidophorid has a suprapreoperculum (see Nybelin, 1966). However, 
Nybelin (1962) has shown in both Leptolepis coryphaenoides and Leptolepis normandica 
a large suborbital, and has indicated that beneath the suborbital of the latter 
there is a separate suprapreoperculum. If this in fact is such a variable feature then 
no importance can be attached to it. But if both suborbital and suprapreoperculum 
are formed from the fragmentation of the preoperculum, as Gardiner (1967a : 199) 
has shown, it is hard to believe that one fragment could completely cover the other, 
especially as the covered fragment is canal bearing. Nevertheless it has already 
been seen that on scale structure the leptolepids are off the ancestry of the salmoni- 
forms (Gardiner, i960). A suprapreoperculum is not known in any other hale- 
costome. It would seem unlikely that further fragmentation of the preoperculum 
within the halecostomes or early teleosts would have taken place since by the 
halecostome grade of organization the preoperculum had become more and more 
intimately associated with both the jaw musculature and the underlying palate 
(Gardiner, 1967a). 

Thus it may be that some as yet unknown pholidophorid which retained the 
parasemionotid suprapreoperculum gave rise to the salmoniforms, gonorhynchi- 



233 



UPPER CRETACEOUS TELEOSTS 



forms and ostariophysans (the basal forms within the Protacanthopterygii) . A 
possible reason for the absence of this ancestral stock and for the apparent absence 
of Salmoniformes before the Lower Cretaceous, and of Ostariophysi before the 



s.pop^ 




Fig. 99. Outline restorations of certain halecostome and teleost skulls to indicate possible 
relationships, (a) Parasemionotus labordei. After Gardiner (i960), (b) Pholidophorus 
bechei. After Nybelin (1966). (c) Leptolepis normandica. After Nybelin (1962). 
(d) Leptolepis coryphaenoides. After Nybelin (1962). (e) Megalops cyprinoides. 
(f) Salvelinus sp. 



ESPECIALLY MYCTOPHOIDS 239 

Tertiary, is that the early evolution of these groups could have occurred in fresh- 
water, and fresh-water deposits are rare. This proposition has already been 
tentatively suggested by Greenwood, et al. (1966 : 347). The Ostariophysi and 
Gonorhynchiformes are predominantly fresh-water forms, while many of the present 
day Salmonoidei spend the breeding part of their lives in fresh-waters. Possibly 
then the ancestral stock of the Protacanthopterygii as a whole was to be found in 
fresh-water. This hypothesis would also help to explain the divergence between the 
Elopiformes and the Leptolepidae on the one hand and the Salmoniformes on the 
other, the initial evolution of the two groups occurring in marine and fresh waters 
respectively. 

No mention in the discussion so far has been given of the Osteoglossomorpha of 
Greenwood, et al. (1966). This division they tentatively considered to have a 
separate origin in the Pholidophoridae. Patterson (1967c : 107) indicates that they 
could have been derived with the Ichthyodectidae from such Jurassic forms as 
Allothrissops. Bardack (1965), however, postulated the derivation of the Chirocen- 
tridae from forms such as Thrissops and Allothrissops. Both Ca vender (1966) and 
Patterson (1967c) refute this, the latter suggesting that these two genera are conver- 
gent with the clupeoids and were derived from a leptolepid and not a pholidophorid 
as Bardack (1965) suggested. Patterson (1967c) thus thinks that the Osteoglosso- 
morpha can be placed with the Elopomorpha, Clupeomorpha and the Leptolepidae 
as a related group of lineages. The osteoglossomorphs however are predominantly 
fresh-water and may have arisen in this medium with some closer affinity with the 
basal protacanthopterygians. 

There is apparently a ' time-lag ' between the origin of the teleosts as a whole and 
their explosive radiation in the Cretaceous. Both the Pholidophoridae and Lepto- 
lepidae occur in the Triassic and the elopiforms are represented in the Jurassic. 
This lag in the radiation of the teleosts during the Jurassic occurred at the time of 
the maximum holostean diversification. Schaeffer (1965) has pointed out that the 
ecological situation at this time is difficult to assess, so that the role that competition 
might have played is impossible to determine. It is possible, as Schaeffer (1965) 
indicated, that the superior halecostome locomotor mechanism plus the plastic 
pholidophoroid jaw design was gradually perfected throughout this period and the 
teleost grade began to exhibit its tremendous potential at the end of the Jurassic. 
It would appear that the elopoids and leptolepids (which arose in the Triassic 
and Lower Jurassic) did not represent a significant advance over the Holostei and 
merely existed as contemporaries. However if the halecostome ancestors of the 
salmoniforms had moved into fresh-water in the late Triassic and Jurassic, and 
gradually evolved during that period in the absence of competition from the holo- 
steans, then this time-lag would not be so apparent. Finally at the close of the 
Jurassic sufficient evolutionary advance had been accomplished by the protacantho- 
pterygians to provide overwhelming competition for the holosteans and the group 
re-entered the seas. Once back in a marine environment in the Cretaceous, the 
Protacanthopterygii rapidly diversified giving the appearance of an ' explosive 
radiation ' at the sub-ordinal and familial levels. 



240 UPPER CRETACEOUS TELEOSTS 

It now remains to consider the evolution of the teleosts within the Cretaceous 
period. The order Elopiformes is represented in the Cenomanian by Sedenhorstia 
which differs little from the Recent Elops and Megalops (Goody, 1969). Already 
Nybelin (1963) has indicated the presence of an elopiform in the Jurassic, thus the 
elopoid assemblage is an ancient one and would appear to have changed little from 
the Jurassic to the present day. Greenwood, et al. (1966 : 355) have indicated that 
the albuloids were established by at least Eocene times and probably earlier since 
they recognize that the Upper Cretaceous Istieus shows strong affinities with Ptero- 
thrissus. The two further derivatives of the basal elopomorphs are the eels and the 
notacanths. Both of these groups are represented in the upper Cretaceous ; eels 
by Urenchelys Woodward (1901) and Anguillavus Hay (1903) ; notacanths by the 
halosaur relatives Echidnocephalus von der Marck (1863) and Enchelurus Woodward 
(1901). These forms are little different from the modern representatives and their 
origin must have been at least in the Lower Cretaceous or even in the Jurassic. 

The basal stock of the Protacanthopterygii probably radiated into marine environ- 
ments at the end of the Jurassic and the beginning of the Cretaceous and is represen- 
ted by the Salmoniformes. Greenwood, el al. (1966) have erected several suborders 
within this order, each representing a small radiation from the basal stock ; the 
Esocoidei, Stomiatoidei, Argentinoidei and Galaxioidei represent certain of these 
radiations. The suborder most closely approaching the basal stock is the Salmon- 
oidei. In the foregoing systematic account four further suborders have been added 
to the Salmoniformes, the Ichthyotringoidei, Cimolichthyoidei, Enchodontoidei and 
Halecoidei, each representing a small radiation of related forms readily separable 
from each other. These fossil suborders represent equivalent groupings to the seven 
or eight salmoniform suborders listed by Greenwood, et al. (1966 : 394). 

The Ichthyotringoidei represents the earliest offshoot from the salmoniform stock 
and is represented in the Gault by Apateodus. This genus existed practically 
unchanged throughout the Upper Cretaceous and finally disappeared in the Maas- 
trichtian. The two families within the suborder, the Ichthyotringidae and the 
Apateopholidae, are both derivable from an Apateodus-like ancestral form by 
extension of the snout (Text-fig. 100). 

The Cimolichthyoidei has been divided into two groups, the cimolichthyid/ 
dercetid group and the prionolepid group. If they are related then the dichotomy 
between the two must have occurred soon after the derivation of the group as a whole 
from the ancestral stock. The Prionolepidae continued into the Turonian without 
diversification. The Dercetidae on the other hand had their origin within the Cimo- 
lichthyidae but radiated quite considerably. For example the very specialized 
long-snouted, long-bodied dercetid Rhynchodercetis extends from the Lower Ceno- 
manian to the Middle Cenomanian and must have arisen in the Lower Cretaceous 
(Albian ?). Pelargorhynchus, although not appearing in the fossil record until 
the Upper Senonian, must have originated from a basal dercetid stock which still 
retained a complete squamation. Since all of the known Upper Cretaceous dercetids 
have a reduced squamation the origin of Pelargorhynchus was presumably well down 
into the Lower Cretaceous, before the origin of the Rhynchodercetis line. Dercetis 



ESPECIALLY MYCTOPHOIDS 



241 



LOWER CRETACEOUS 



/ ,.- T Halec 



, -Enchodus 



CENOMANIAN 
Hemisaurida 



"Phylactocephalus 



"-— Eurypholis- 
Saurorhamphus 

•Rhynchodercetis 
,-Dercetis — 



"Apateodus- 



^Prionolepis- 



Ichthyotringa- 



TURONIAN 



-Cimolichthys- 



SENONIAN 



~"~Palaeolycus 



MAAS. 



-Pelargorhynchus 



""-Apateopholis 



Fig. 100. Evolutionary tree of the fossil salmoniform suborders. 

itself extends through the Upper Cretaceous and disappears at the end of the 
Senonian (Text -fig. 100). 

The Enchodontoidei is composed of two families, the Enchodontidae and Eury- 
pholidae. The genus Enchodus extended throughout the Upper Cretaceous and is 
recorded from the Tertiary (Arambourg, 1952). The divergence between the two 
families must have occurred in the Lower Cretaceous since a specialized derivative 
of the Eurypholidae, Saurorhamphns, is present in the lowermost Cenomanian. The 
genus Eurypholis is not recorded until the Middle Cenomanian and extends through 
into the Turonian. Saurorhamphus can be derived directly from Eurypholis, merely 
by slight snout and body extensions. Palaeolycus represents a specialized derivative 
of the Enchodontidae and occurs in the Upper Senonian. The origin of this genus 
can be visualized as being from one of the shallower bodied species of Enchodus, in 
which vertebral multiplication occurred. This origin was presumably in the Upper 



242 UPPER CRETACEOUS TELEOSTS 

Cretaceous, since there are no factors excluding the genus Enchodus from having been 
the immediate ancestor of Palaeolycus (Text-fig. ioo). 

The Halecoidei are represented in the Lower Cenomanian by the most primitive 
genus Halec (d'Erasmo, 1946). Phylactocephalus occurs in the Middle Cenomanian 
and could easily have been derived directly from Halec at the end of the Lower 
Cretaceous. Hemisaurida also had its origins within the same basic stock, but would 
appear to have branched off at a slightly earlier stage than Phylactocephalus, since 
Hemisaurida is already represented in the Lower Cenomanian. The genus Halec 
remained the most generalized form and extended into the Senonian (Text-fig. 100) . 

In the discussions of these fossil salmoniforms, I have attempted to show that they 
have evolved along lines represented throughout the division Protacantho- 
pterygii and to a more marked extent parallel the Myctophiformes and the 
Acanthopterygii. Each of the suborders dealt with earlier appears to express similar 
evolutionary capabilities. Within the recent Salmoniformes the Galaxioidei 
(Galaxias and Aplochiton) are often cited as being practically indistinguishable from 
the Myctophiformes especially with regard to the upper jaw (Greenwood, et at., 
1966 : 366 ; Gosline, Marshall and Mead, 1966 : 2). Thus a recent suborder 
reinforces the suggestion of the similarity in evolutionary potential between the 
fossil salmoniforms and the myctophiforms. It would seem that several of the 
salmoniform suborders have the potential, latent or expressed, to produce a more 
advanced structural complex which in most cases approaches the myctophiform 
grade. Thus the myctophiform grade (although it is proposed in this work that the 
order Myctophiformes is itself monophyletic) can be attained by other Salmoni- 
formes. 

The acanthopterygian complex of characters represents a considerable advance 
over and above the most advanced Salmoniformes and Myctophiformes. The 
major advances in feeding, locomotion and protection are exemplified by the 
protrusile upper jaw, thoracic pel vies, body shortening and deepening, and the 
development of spines in fins and on scales. These features are only encountered 
together in the single line stemming from the protacanthopterygian order Cteno- 
thrissiformes (Patterson 1964). (True protrusibility of the jaws never occurs in the 
Paracanthopterygii, although all of the other features listed above are shown.) The 
Ctenothrissiformes have been shown to be little removed from a basal salmoniform 
stock (Table I, p. 203) and also to be very similar to a basal myctophiform (e.g. 
Sardinioides). The common ancestry of these two groups (Ctenothrissiformes and 
Myctophiformes) occurs in the Lower Cretaceous. The Myctophiformes are well 
established by the Middle Cenomanian and would appear to have had their origin in 
the Albian. Nematonotus has been mentioned as being a genus which approaches 
both Myctophiformes and Ctenothrissiformes, substantiating the basal convergence 
of the two groups. Nematonotus occurs in the Cenomanian together with Acro- 
gnathus, which also approaches both groups. Sardinius, however is a probable 
derivative of the very early myctophiform lineage despite its late appearance in the 
fossil record (Upper Senonian) . Sardinius closely parallels the family Myctophidae, 
particularly in the backwardly inclined suspensorium and the absence of supra- 



ESPECIALLY MYCTOPHOIDS 



243 



LOWER CRETACEOUS 



CENOMANIAN 



TURONIAN 



^Pattersonichthys 



-Ctenothrissa- 



Nematonotus 
Acrognathus- 



Sardinioides- 



Volcichthys 



SENONIAN 



—- Opisthopteryx 
--Sardinius 



»Paracanthopterygii 



^Myctophiformes 



Fig. ioi. 



Evolutionary tree of the Myctophiformes and Ctenothrissiformes within 
the Cretaceous. 



maxillae. Volcichthys, from the Lower Cenomanian, is tentatively suggested to be 
related to Sardinius. The Upper Senonian genus Opisthopteryx appears to have 
been derived from a form not unlike Sardinioides (Text-fig. 101). 

Patterson (1967a) has indicated that the Beryciformes probably arose sometime 
in the Albian from a ctenothrissiform stock. By the end of the Lower Cretaceous 
salmoniform derivatives were numerous and had undergone a considerable degree of 
radiation. Having given rise to the Ctenothrissiformes and the Myctophiformes, 
the basal stock appears to have continued and produced yet further lineages, none 
however corresponding to the ctenothrissiforms but some paralleling the myctophi- 
forms. The Stomiatoidei is possibly represented in the Lower Cenomanian by 
Paravinciguerria (Arambourg, 1954). Weitzman (1967 : 535-537) has evaluated the 
fossil record of the stomiatoids, and has reached the conclusion that Paravinciguerria 
may be stomiatoid, whereas Idrissia (Arambourg, 1954) is probably not. The 
Argentinoidei, Galaxioidei and Esocoidei do not appear in the fossil record until the 
Tertiary (Patterson, 1967b). The argentinoids and galaxioids presumably arose in 
the Upper Cretaceous or actually in the Tertiary itself, but the Esocoidei, in view 
of the structure of their dermal upper jaw, present more of a problem. The separate 
proethmoid of Esox and Dallia may be the forerunner of an ascending process or it 
might be only an analogous structure indicating the derivation of the Esocoidei 



244 UPPER CRETACEOUS TELEOSTS 

from the basal salmoniform stock before forms such as the myctophiforms and 
ctenothrissiforms had arisen. Weitzman (1967 : 526) doubts that the proethmoids 
of esocoids have any relation to the evolution of the ascending premaxillary processes 
in other teleosts. He suggests that the esocoids exhibit an experiment in jaw 
design which did not impart any significant evolutionary advantages. Due to 
the fresh-water habitat of the Esocoidei this group may represent an early offshoot 
from the salmoniform ancestral stock in the Jurassic fresh waters. 

The fossil salmoniform suborders Enchodontoidei, Ichthyotringoidei, Cimolich- 
thyoidei and Halecoidei extend throughout the greater part of the Upper Cretaceous, 
but all seem to have disappeared by the Maastrichtian. The myctophiforms are 
represented throughout the Upper Cretaceous but the ctenothrissiforms had 
disappeared by the Middle Turonian, this latter group having been replaced by its 
more advanced descendants, the Beryciformes. The Beryciformes were evolving 
and diversifying throughout the Cretaceous, and towards the close of this period 
several of the lines represented in the present day Beryciformes and Perciformes 
were established (Patterson, 1964, 1967a, 1967b). 

In the Cretaceous period the Elopiformes, Anguilliformes, Notacanthiformes, 
Clupeiformes, Osteoglossiformes, Salmoniformes, Myctophiformes, Ctenothrissi- 
formes and Beryciformes would have inhabited similar environments and would have 
been in direct competition with each other for food and living space. The Cretaceous 
fauna would have lived in the most productive parts of the ocean, these favourable 
parts being inshore waters (above the upper reaches of the continental shelves) or 
the surface waters of equatorial oceanic regions (Marshall, 1963 : 189). Throughout 
the Upper Cretaceous the Beryciformes were increasing and perfecting the protrusile 
mouth which opened up new and hitherto unused food sources. At the same time 
the Beryciformes were becoming more effective swimmers and were developing 
protective spines on fins and scales. A true protrusile mouth is not developed in any 
other group, although all the basic requirements are present in the Myctophiformes. 
Towards the close of the Cretaceous the Beryciformes gave rise to lines leading on to 
the Perciformes and other advanced Acanthopterygii (Patterson, 1964, 1967a). 
Thus the beryciforms and their derivatives would have provided stern competition 
for the less labile, more primitive grades of organization represented by the Elopi- 
formes and the early salmoniform derivatives. Many of these lower teleostean 
groups succumbed to this pressure of competition and died out, e.g. Enchodontoidei, 
Cimolichthyoidei, etc. Others, better adapted to their particular habits, e.g. 
Clupeomorpha, or modified along very specialized lines, e.g. eels, were able to survive 
and compete. Still others at the end of the Cretaceous were forced to seek living 
space elsewhere. This involved the colonization of the less productive parts of the 
ocean, which include the central water masses at increasing depths. Thus groups 
like the myctophiforms and stomiatoids were forced into deeper waters. It is 
interesting to note that the vast majority of the recent bathypelagic fish are all lower 
teleostean in origin. Once in the bathypelagic environment a certain amount of 
adaptive radiation has occurred in both of the above mentioned groups, this radiation 
commencing at the close of the Cretaceous and the beginning of the Tertiary. Thus 



ESPECIALLY MYCTOPHOIDS 



245 




Fig. 102. Evolutionary tree of the Teleostei from the Triassic to the present day. 



246 UPPER CRETACEOUS TELEOSTS 

the origin of a considerable proportion of the recent bathypelagic fauna is accounted 
for when viewed in the light of the competition for food and living space with the 
Acanthopterygii. 

The evolutionary considerations outlined in the present section are summarized 
in Text-figure 102. 

V. CONCLUSIONS 

In the foregoing descriptions and discussion some mention has been made of the 
divergence between on the one hand the Elopomorpha and on the other the salmoni- 
form group and its derivatives. The Elopiformes have been considered in the last 
section to be very close to the halecostomes, with the Elopoidei retaining the greatest 
number of halecostome characters (Nybelin, 1956). Certain specializations, however, 
rule this group out from the main course of teleost evolution (Greenwood, et al., 
1966 : 348). The Salmoniformes do not possess any of the limitations of the 
elopiforms and it has been shown that the major teleost radiations had their origin 
within the Salmoniformes. 

The reality of the divergence between the Elopiformes and the Salmoniformes has 
been indicated but this divergence may in part be due to the absence of fossil 
salmoniforms in deposits earlier than the Albian. The elopiform lineage is distinct 
by the Upper Jurassic and undoubtedly arose somewhat earlier. The salmoniforms, 
first represented in the marine Albian, may have also been present in the Jurassic, 
but in fresh water. If this is the case then the basal members of the two groups 
would possibly have shown a much closer resemblance. The Ichthyotringoidei has 
been proposed as being the most primitive of the Cretaceous salmoniforms so far 
described and certain osteological features, notably the caudal skeleton, are as 
easily comparable with an elopiform as with a salmoniform. Although the two 
lineages might approach one another more closely basally, both probably having been 
derived from the Pholidophoridae, there appears to have been a marked divergence 
in evolutionary potential between the two groups. 

The structure of the jaws has been stressed as being one of the major factors 
influencing the course of later teleost evolution. In the Elopomorpha the jaw is 
non-protrusile and there is no mechanism enabling the premaxilla to be moved 
forwards, while the maxillae form the majority of the upper jaw margin. The 
salmoniform derivatives on the other hand exhibit protrusibility of several types, the 
essential factor being that the jaw is capable of becoming protrusile. Eaton (1935 : 
168) and Alexander (1966 ; 1967a ; 1967b : 250) have outlined three main types 
of jaw protrusion, an ostariophysan type, a cyprinodont type and a true acantho- 
pterygian type. Eaton goes on to say that these three types possess an enlarged 
premaxilla with a median dorsal process which suggests a common origin. Con- 
sidering the comments of Allis (1909), Greenwood, et al. (1966), Weitzman (1967) 
and Goody (1968) it is possible that the dorsal processes are homologous and the 
three types share a common ancestry. 

Supporting evidence for the closeness of the basal stocks of the Salmoniformes, 
Myctophiformes, Ctenothrissiformes and Beryciformes is afforded by the structure 



ESPECIALLY MYCTOPHOIDS 247 

of the caudal skeleton. This complex is remarkably similar in all four groups and 
undergoes identical fusions and reductions. 

Other trends exhibited by the salmoniform derivatives also separate them from 
the elopiforms, for example the constellation of characters associated with the paired 
fins and increased manoeuvrability. However fin spines are produced in the more 
specialized derivatives of both lineages (Marshall, 1962, has demonstrated the 
presence of true fin spines in the Heteromi) being due to the stresses imposed on the 
fins by increased momentum and bodily flexion (Patterson, 1964). 

Thus after the separation of the elopiform and salmoniform lineages at the level 
of the pholidophorid halecostomes the salmoniforms appear to have evolved possibly 
in fresh water for some considerable time. During this period many changes were 
effected within the Salmoniformes. Their subsequent radiation back into marine 
environments in the Lower Cretaceous involved much ' experimentation ', both 
successful and unsuccessful, in respect of survival to the present day. These 
evolutionary ' experiments ' have all been along somewhat similar lines with the 
same changes and deletions occurring time and again. These changes have been 
manifested in all parts of the skeleton, the skull, vertebral column, paired fins and 
girdles, and lastly in the caudal skeleton. However the majority of groups have 
also exhibited changes and specializations peculiar to themselves as well as showing 
the general trends present in all of the groups. Thus a variety of radiations occurred 
into particular ecological niches. Certain of these radiations are represented by the 
salmoniform suborders Cimolichthyoidei, Enchodontoidei and Halecoidei. In many 
groups the specialization for a particular mode of life has proceeded to such an extent 
that often any evolutionary plasticity exhibited by the parent stock is lost. Whilst 
filling their ecological niche successfully (success being based on their relative 
abundance in the fossil record) these forms seem to have been adversely affected by 
competition from later radiations. These later forms have advanced to a greater 
degree of efficiency in many aspects of life but still seem to have retained a more 
generalized overall construction and are still capable of further variability. This 
competitive aspect presumably accounted for much of the extinction of groups 
successful through the Upper Cretaceous. 

VI. ACKNOWLEDGEMENTS 

I wish to thank Dr. E. I. White, F.R.S., of the British Museum (Natural History) 
for the opportunity of studying the collections which form the basis of this work. I 
am most grateful to Dr. B. G. Gardiner of Queen Elizabeth College (London Univer- 
sity) for the supervision of my work and for much valuable advice ; to Professor G. 
Chapman in whose department the work was done ; to Dr. P. H. Greenwood and 
Dr. C. Patterson of the British Museum (Natural History) for much fruitful discus- 
sion ; and to Mr. H. A. Toombs and Mr. C. I. Macadie also of the British Museum for 
their generous help. Grateful acknowledgement is also made to the Central Research 
Fund of London University for a grant enabling me to visit many European Museums 
and University departments. Finally I extend my thanks to Mrs. A. E. Jordan of 
Queen Elizabeth College (London University) for valuable help in photographic and 
preparatory techniques. 



UPPER CRETACEOUS TELEOSTS 



VII. REFERENCES 



Agassiz, L. 1833-44. JRecherches sur les poissons fossiles, 5 vols., 1420 pp., 396 pis., with 
supplement. Neuchatel. (For dates of publication see Woodward and Sherborn, 1890.) 

Alexander, R. McN. 1966. The functions and mechanisms of the protrusible upper jaws of 
two species of cyprinid fish. J . Zool., Lond., 149 : 288-296, 6 pis. 

1967a. The functions and mechanisms of the protrusible upper jaws of some acanthop- 

terygian fish. J. Zool., Lond., 151 : 43-64, 6 pis. 

1967b. Mechanisms of the jaws of some atheriniform fish. J. Zool., Lond., 151 : 233- 

255. 1 pl- 
Allis, E. P. 1909. The cranial anatomy of the mail-cheeked fishes. Zoologica, Stuttgart, 57, 

iii + 219 pp., 8 pis. 
Arambourg, C. 1925. Revision des poissons fossiles de Licata (Sicile). Annls Paleont., 

Paris, 14 : 39-132, 10 pis., 12 figs. 
1944- Note preliminaire sur quelques poissons fossiles nouveaux. Bull. Soc. gdol. Fr., 

Paris (5), 8 : 281-285, 1 pl. 
1952. Les Vertebres fossiles des gisements de Phosphates. (Maroc, Algerie, Tunisie.) 

Notes Mdm. Sew. Mines Carte, geol. Maroc, Rabat, 92 : 1-372, 44 pis. 
1954- Les poissons cretaces du Jebel Tselfat. Notes M6m. Serv. Mines Carte geol. 

Maroc, Rabat, 118 : 1-188, 20 pis., 68 figs. 
Bardack, D. 1965. Anatomy and evolution of chirocentrid fishes. Paleont. Contr. Univ. 

Kansas, Lawrence, Vertebrata, 10 : 1-88, 2 pis., 27 figs. 
Barrois, C. 1874. Catalogue des poissons fossiles du Terrain Cretace du Nord de la France. 

Bull, scient. hist. lift. Nord., Lille, 6 : 101-110 and 130-136. 
Bassani, F. 1879. Vorlaufige mitteilungen liber die (fossile) fischfauna der Insel Lesina. 

Verh. K. K. geol. Reichsanst., Wien, 8 : 161-168. 
1882. Descrizione dei pesci fossile di Lesina, accompagnata da appunti su alcune altre 

ittiofaune cretacee. Denkschr. Akad. Wiss., Wien, 56, 1 : 195-288, 16 pis. 
Beebe, W. & Crane, J. 1939. Deep-sea fishes of the Bermuda oceanographic expeditions. 

Family Melanostomiatidae. Zoologica, New York, 24, (2), 6 : 65-238. 
de Beer, G. R. 1954- Archaeopteryx and evolution. Advmt. Sci., Lond., 11 : 160-170. 
Berg, L. S. 1940. Classification of fishes both Recent and Fossil (First edition, Russian with 

complete English translation). Trudy zool. Inst., Leningr., 5, 2 : 517 pp., 190 figs. 
1955- Classification of fishes and fish-like vertebrates, living and fossil (Second edition, 

corrected and enlarged). Trudy zool. Inst. Leningr., 20 : 1-286, 263 figs. 
Bertin, L. & Arambourg, C. 1958. Superordre des Teleosteens : 2204-2500, figs. 1561- 

1788. In Traitd de Zoologie, 13, ed. Grasse, P-P. Paris : Masson et Cie. 
Cavender, T. 1966. The caudal skeleton of the Cretaceous teleosts Xiphactinus, Ichthy- 

odectes, and Gillicus, and its bearing on their relationship with Chirocentrus. Occ Pap. 

Mus. Zool. Univ. Mich., Ann Arbor, 650 : 1-15, 1 pl. 
Cope, E. D. 1870. On the Saurodontidae. Proc Am. phil. Soc, Philadelphia, 11 : 529-538. 
1878. Description of fishes from the Cretaceous and Tertiary deposits West of the 

Mississippi River. Bull. U.S. geol. geogr. Surv. Territ., 4, 1 : 67-77. 
Crane, J. M. 1966. Late Tertiary radiation of viperfishes (Chauliodontidae) based on a 

comparison of Recent and Miocene species. Sci. Contr. Los Angeles County Mus., 115: 

1-29, 13 figs. 
Dames, W. 1887. Ueber die gattung Saurodon Hays. Sber. Ges. natnrf. Freunde Berl., 

5 : 72-78. 
Danil'chenko, P. T. 1964. Superorder Teleostei : 396-472, pis. 5-14, in Osnovy Palaeon- 

tologii, 11, Beschelyustnye ryby. ed. Obruchev, D. B., Akad. Nauk SSSR. (In Russian). 
Dartevelle, E. & Casier, E. 1941. Les poissons fossiles de 1' Angola. Comuncoes Servs. 

geol. Port., Lisboa, 22 : 99-109, 1 pl. 



ESPECIALLY MYCTOPHOIDS 249 

Dartevelle, E. & Casier, E. 1949. Les poissons fossiles du Bas Congo et des regions 

voisines (2me partie). Annls Mus. r. Congo beige Ser. 4 to. Min., giol., paleont., Tervuren, 

(3), 2, 2 : 201-256, pis. 17-22, 16 figs. 
Davis, J. W. 1887. The fossil fishes of the Chalk of Mount Lebanon in Syria. Scient. Trans. 

R. Dubl. Soc, 3, (2) : 457-636, 25 pis. 
Dixon, F. 1850. The geology and fossils of the Tertiary and Cretaceous formations of Sussex. 

422 pp., 40 pis. London : F. Dixon. 

1878. Ibid, (second edition). 

Dunkle, D. H. 1940. The cranial osteology of Notelops brama (Agassiz), an elopid fish from 
the Cretaceous of Brazil. Lloydia, 3 : 157-190, 9 figs. 

Eaton, T. H. 1935. Evolution of the upper jaw mechanism in teleost fishes. /. Morph., 

Philadelphia, 58, 1 : 157-172, 2 pis., 15 figs. 
Egerton, P. M. G. 1850. In Dixon 1850. {see above). 
d'Erasmo, G. 191 2. 11 Saurorhamphus freyeri Heckel degli scisti bituminosi cretacei del 

Carso Triestino. Boll. Soc. adriat. Set. nat., Trieste, 26, 1 : 45-88, 2 pis., 15 figs. 

1946. LTttiofauna cretacea dei dintorni di Comeno nel Carso Triestino. Atti Accad. 

Sci.fis. mat., Napoli, (3a), 2, 8 : 136 pp., 1 pi. 34 figs. 

1952. Nuovi ittioliti cretacei del Carso Triestino. Atti Mus. civ. Slor. nat. Trieste, 18, 

4 : 81-122, 3 pis., 12 figs. 

Forir, H. 1887. Contributions a l'etude du systeme cretace de la Belgique. 1. Sur quel- 
ques poissons et crustaces nouveaux ou peu connue. Annls Soc. gSol. Beige., Liege, 14 : 
25-56, 2 pis. 

Gardiner, B. G. i960. A revision of certain actinopterygian and coelacanth fishes, chiefly 
from the Lower Lias. Bull. Br. Mus. nat. Hist. (Geol.), London, 4, 7: 239-384, pis. 36-43. 

1967a. The significance of the preoperculum in actinopterygian evolution. J. Linn. 

Soc. Zoology, London, 47, 311 : 197-209, 8 figs. 

1967b. Chondrostei, Holostei and Halecostomi, in Harland, W. B. et al. The fossil record. 

Lond. Geological Society. Pp. 644-654. 
Geinitz, H. B. 1875. Das Elbthalgebirge in Sachsen. Zweiter theil. Der mittlere und 

obere Quader VI. Wurmer, Krebse, Fische, Reptilien und Pflanzen. Palaeontographica, 

Cassel, 20, 2 : 200-245, 10 pis. 
Giebel, C. G. A. 1848. Fauna der Vorwelt mil steter Berucksichtigung der lebenden Thiere. 

1, 3, Die Fische, xii + 467 pp., Leipzig. 
Goody, P. C. 1968. The skull of Enchodus faujasi from the Maastricht of Southern Holland. 

Proc. Kon. Nederl. Acad. Wetensch., Amsterdam, Ser. B., 71 : 209-231, 9 figs. 

1969. Sedenhorstia dayi (Hay), a new elopoid from the Cenomanian of Hajula in the 

Lebanon. Am. Mus. Novit., New York, 2358 : 1-23, 9 figs. 

Gosline, W. A. i960. Contributions toward a classification of modern isospondylous fishes. 

Bull. Br. Mus. nat. Hist. (Zool.), London, 6 : 325-365, 15 figs. 
■ 1961. Some osteological features of modern lower teleostean fishes. Smithson. Misc. 

Coll., Washington, 142, 3 : 1-42, 8 figs. 

1963. Considerations regarding the relationships of the percopsiform, cyprinodontiform 

and gadiform fishes. Occ. Pap. Mus. Zool. Univ. Mich., Ann. Arbor, 629 : 1-38, 11 figs. 

Gosline, W. A., Marshall, N. B. & Mead, G. W. 1966. Iniomi : Characters and Synopsis 

of Families. Mem. Sears Fdn. mar. Res., New Haven, 1, 5 : 1-18, figs. 1-6. 
Gosselet, J. 1883. Esquisse geologique du Nord de la France et des Contrees voisines. 

Archs. Soc. geol. N ., Lille, 3 : 279-342, pis. 22-29. 
Greenwood, P. H., Rosen, D. E., Weitzman, S. H. & Myers, G. S. 1966. Phyletic studies 

of teleostean fishes, with a provisional classification of living forms. Bull. Am. Mus. nat. 

Hist., New York, 131 : 339-456, pis. 21-23, 3 2 charts. 



2 5 o UPPER CRETACEOUS TELEOSTS 

Gregory, W. K. 1933- Fish skulls: a study of the evolution of natural mechanisms. Trans. 

Am. phil. Soc, Philadelphia, 23, n : 75-481, 299 figs. 
Griffith, J. & Patterson, C. 1963. The structure and relationships of the Jurassic fish 

Ichthyokentema puvbeckensis. Bull. Br. Mus. nat. Hist. (Geol.), London, 8, 1 : 1-43, pis. 

1-4. 
Gunther, A. 1864. Description of a new fossil fish from the Lower Chalk. Geol. Mag., 

London, 1, 3 : 114-118, pi. 6. 
Harris, J. E. 1938. The role of the fins in the equilibrium of the swimming fish. II. The 

role of the pelvic fins. /. exp. Biol., Cambridge, 15 : 32-47, 8 figs. 
Harry, R. R. 1953. Studies on the bathypelagic fishes of the family Paralepididae. 1. 

Survey of the genera. Pacif. Sci., Honolulu, 7, 2 : 219-249. 
Hay, O. P. 1903. On a collection of Upper Cretaceous fishes from Mount Lebanon, Syria, 

with description of four new genera and nineteen new species. Bull. Am. Mus. nat. Hist., 

New York, 19, 10 : 395-452, pis. 24-37. 
Hebert, E. 1855. Tableau des fossiles de la Craie de Meudon, et description de quelques 

especes nouvelles. Mim. Soc. geol. Fr., Paris, 5, 4 : 345-356. 1 pi., 10 figs. 
Heckel, J. J. 1849. Die fossilen fische des Libanon, 2, 3 : 336-346, pi. 23, of Reisen in 

Europa, Asien und Afrika, ed. von Russeger, J. Stuttgart : E. Schweizerbart. 

1850. Beitrage zur Kenntniss der fossilen fische Oesterreiche. Denkschr. Akad. Wiss., 

Wien, 1 : 201-243, 15 pis. 

Jordan, D. S. 1905. A guide to the study of fishes, 2, xxii + 589 pp., 506 figs. New York : 
H. Holt and Co. 

1923. A classification of fishes, including families and genera as far as known. Stanford 

Univ. Pubis., Biol. Sci., Palo Alto, 3, 2 : 79-243. 

Kner, R. 1867. Neuer beitrag zur Kenntniss der fossilen fische von Comen bei Gorz. Sber. 

Akad. Wiss. Wien, (1), 56 : 171-200, 5 pis. 
Kramberger, D. G. 1881. Studien fiber die gattung Saurocephalus Harlan ; ein beitrag 

zur Neocom-Fischfauna der Insel Lesina. Jb. geol. Reichsanst, Wien, 31 : 371-379. 

1895. De piscibus fossilibus Comeni, Mrzleci, Lesinae et M. Libanonis. Djela jugosl. 

Akad. Znan. Umjeln., Zagreb, 16 : 1-67, 12 pis. 

Kruizinga, P. 1924. Apateodus corneti (For.) in the Senonian beds of the Southern part of 
Limburg (Netherlands). Proc. Sect. Sci. K. ned. Akad. Wet., Amsterdam, 27 : 293-312, 
2 pis. 

1952. Twee nieuwe Koppen van Apateodus corneti (For.). Natuurh. Maandbl, Maastricht, 

41 : 42-46, 5 figs. 

Lehman, J. P. 1966. Actinopterygii : 1-142, figs. 1-211, in Traiti de Paleontologie, 4, 3, ed. 

Piveteau, J. Paris, Masson et Cie. 
Leidy, J. 1857. Remarks on Saurocephalus and its allies. Proc. Acad. nat. Sci. Philad., 

8 : 301-302. 
Leonardi, A. 1966. L'ittiofauna Cenomaniana di Floresta — Messina. Palaeontographia 

Italica, Pisa, 60, (1965) : 33-67, pis. 10-15. 
Leonhard, R. 1897. Die Fauna der Kreideformation in Oberschlesein. Palaeontographica, 

Stuttgart, 44 : 1 1-70, 4 pis. 
Leriche, M. 1902. Revision de la faune ichthyologique des terrains crdtaces du Nord de la 

France. Annls Soc. giol. N ., Lille, 31 : 87-146, pis. 2-4. 
— ■ — 1906. Contribution a l'etude des poissons fossile du Nord de la France et des regions 

voisines. Mim. Soc. gtol. N., Lille, 5, 1 : 1-430, 17 pis. 
1929. Les poissons Cretace marin de la Belgique et du Limbourg hollandais (note pr6- 

liminaire). Bull. Soc. beige Giol. Paliont. Hydrol., Bruxelles, 37 : 199-299, 19 figs. 
Loomis, F. 1900. Die Anatomie und die Verwandtschaft der Ganoid und Knochen Fische aus 

der Kreide Formation von Kansas. Palaeontographica, Stuttgart, 46 : 213-284, pis. 19-27. 
von der Marck, W. i 858. Ueber einige Wirbelthiere, Kruster und Cephalopoden der 

Westphalischen Kreide. Z. dt. geol. Ges., Berlin, 10 : 231-271. 



ESPECIALLY MYCTOPHOIDS 251 

von der Marck, W. i 863. Fossile Fische, Krebse und Pflanzen aus dem Plattenkalk der 
Jungsten Kreide in Westphalen. Palaeontographica, Cassel, 11 : 1-83, 14 pis. 

1873. Neue beitrage zur Kenntniss der fossilen fische und anderer Thierreste aus der 

jungsten Kreide Westfalens. Palaeontographica, Cassel, 22 : 55-74, 2 pis. 

1885. Fische der Oberen Kreide Westfalens. Palaeontographica, Cassel, 31 : 233-267, 

5 pis. 

Mantell, G. A. 1822. The fossils of the South Downs; or Illustrations of the Geology of Sussex. 

320 pp., 42 pis. London : Lupton Relfe. 
Marshall, N. B. 1955. Studies of alepisauroid fishes. ' Discovery ' Rep., 27 : 303-336, 

pi. 19. 

1961. A young Macristium and the Ctenothrissid fishes. Bull. Br. Mus. nat. Hist. 

(Zool.), London, 7, 8 : 353-370, 4 figs. 

1962. Observations on the Heteromi, an order of teleost fishes. Bull. Br. Mus. nat. 

Hist. (Zool.), London, 9, 6 : 249-270, 5 figs. 

1963. Diversity, distribution and speciation of deep-sea fishes. Syst. Assoc, 5 : 181- 

195, 2 figs. 

Maury, C. J. 1930. O creataceo da Parahyba do Norte. Monografias Div. geol. miner. Bras., 

Rio de Janeiro, 8 : 4-305, 35 pis. 
Monod, T. 1959. Notes sur l'epine latero-caudale et la queue de YAcanthurus monroviae. 

Bull. Inst. fr. Afr. noire, Paris, (A), 21, 2 : 710-734, 32 figs. 
1967. Le complexe urophore des Teleosteens : typologie et evolution (note preliminaire) . 

Collogues int. Cent. natn. Rech. scient., Paris, 163 : 111-131, 16 figs. 
Nelson, G. J. 1967. Branchial muscles in some generalized Teleostean fishes. Acta zool., 

Stockholm, 48 : 277-288, 2 figs. 
Newton, E. T. 1878. Remarks on Saurocephalus, and on the species which have been 

referred to that genus. Q. J I geol. Soc. Lond., 34 : 786-796. 
Norden, C. R. 1961. Comparative osteology of representative Salmonid Fishes, with 

particular reference to the grayling (Thymallus arcticus) and its phylogeny. /. Fish. Res. 

Bd. Can., Ottawa, 18, 5 : 679-791, 16 pis., 2 figs. 
Nursall, J. R. 1956. The lateral musculature and the swimming of fish. Proc. zool. Soc. 

Lond., 126 : 127-143. 
Nybelin, O. 1956. Les canaux sensoriels du museau chez Elops saurus. Ark. Zool., Stock- 
holm, (2), 10, 9 : 453-458, 3 figs. 

1962. Preliminary note on two species previously named Leptolepis bronni Agassiz. 

Ark. Zool., Stockholm, (2), 15, 18 : 303-306, 1 fig. 

1963. Zur morphologie und terminologie des Schwanzskelettes der Actinopterygier. 

Ark. Zool., Stockholm, (2), 15, 35 : 485-516, 22 figs. 

1966. On certain Triassic and Liassic representatives of the family Pholidophoridae s. str. 

Bull. Br. Mus. nat. Hist. (Geol.), London, 11, 8 : 351-432, 15 pis., 16 figs. 

1967. Notes on the reduction of the sensory canal system and of the canal-bearing bones 

in the snout of higher actinopterygian fishes. Ark. Zool., Stockholm, (2), 19, 9 : 235-246, 

4 figs. 
Patterson, C. 1964. A review of Mesozoic acanthopterygian fishes with special reference to 

those of the English Chalk. Phil. Trans. R. Soc, London (B), 247 : 213-482, pis. 2-5. 
1967a. New Cretaceous berycoid fishes from the Lebanon. Bull. Br. Mus. nat. Hist. 

(Geol.), London, 14, 3 : 69-109, 4 pis., 11 figs. 

1967b. Teleostei, in Harland, W. B. et al. The fossil record. Lond. Geological Society. 

Pp. 654-666. 

1968a. The caudal skeleton in Lower Liassic Pholidophorid fishes. Bull. Br. Mus. nat. 

Hist. (Geol.), London, 16, 5 : 201-239, 5 pis. 



252 UPPER CRETACEOUS TELEOSTS 

Patterson, C. 1968b. The caudal skeleton in Mesozoic acanthopterygian fishes. Bull. 

Br. Mus. nat. Hist. (Geol.), London, 17, 2 : 47-102, 28 figs. 
Pictet, F. J. 1850. Description de quelques poissons fossiles du Mont Liban, 59 pp., 10 pis., 

Geneve : J-G. Fick. 
& Humbert, A. 1866. Nonvelles recherches sur les poissons fossiles du Mont Liban, 

vii + 114 pp., 19 pis. Geneve : Georg. 
Priem, F. 1908. Etudes des poissons fossiles du Bassein Parisien. Annls Paleont., Paris 

3 : 1-144. 5 pis., 74 fi gs- 
Kavner, D. H. 1937. On Leptolepis bronni Agassiz. Ann. Mag. nat. Hist., London, (10) 

19 : 46-74, 14 figs. 

1948. The structure of certain Jurassic holostean fishes, with special reference to their 

neurocrania. Phil. Trans. R. Soc., London (B), 233, : 287-345, 4 pis., 35 figs. 

Regan, C. T. 1909. The classification of teleostean fishes. Ann. Mag. nat. Hist., London, 
(8) 3 : 75-86. 

- 191 1. The anatomy and classification of the teleostean fishes of the order Iniomi. Ann. 
Mag. nat. Hist., London, (8) 7 : 120-133, 7 fig s - 

& Trewavas, E. 1930. The fishes of the families Stomiatidae and Malacosteidae. Rep. 

Dana Exped., Copenhagen, 6 : 1-143, J 4 pis., I 3^ figs. 
Reuss, A. E. 1845. Die versteinerungen der bohmischen Kreide formation. Stuttgart, 1 : 

158 pp., pis. 1-13. 
Rixon, A. E. 1949. The use of acetic and formic acids in the preparation of fossil vertebrates. 

Museums J ., London, 49 : 116-117. 
Rode, K. 1937- Uber Enchodus in der Oberen Kreide von Quedlinburg. Palaont. Z., Berlin, 

19 : 127-133, pi. 9. 
Romer, A. S. 1966. Vertebrate Palaeontology, 3rd edition. 468 pp., 443 figs., Chicago University 

Press. 
Rosen, D. E. 1962. Comments on the relationships of the North American cave fishes of the 

family Amblyopsidae. Am. Mus. Novit., New York, 2109 : 1-35, 24 figs. 

1964. The relationships and taxonomic position of the half beaks, killifishes, silversides, 

and their relatives. Bull. Am. Mus. nat. Hist., New York, 127, 5 : 217-268, 2 pis. 

Saint-Seine, P. de. 1949. Les poissons des Calcaires Lithographiques de Cerin (Ain). Nouv. 
Archs. Mus. Hist. nat. Lyon, 2: vii 4- 357 PP-, 26 pis. 

Sarra, R. 1933. Denti di Pesci del Cretaceo e di Mammiferi del Pliocene rinvenuti in Basili- 
cata. Riv. ital. Paleont. Stratigr., Parma, 39 : 29-34. 

Sauvage, H. E. 1883. Notes sur les poissons fossiles — Sur l'intermaxillaire de YEnchodus 
lewesiensis Mantell. Bull. Soc. geol. Fr., Paris, (3) 11 : 480-481, pi. 12, fig. 16. 

Schaeffer, B. 1965. The role of experimentation in the origin of higher levels of organiza- 
tion. Sysl. Zool., Washington, 14 : 318-336, 9 figs. 

Siegfried, P. 1954. Die Fisch-Fauna des Westfalischen Ober-Senons. Palaeontographica, 
Stuttgart, (A), 106, 1 : 1-36, 15 pis., 2 figs. 

- 1966. Zur osteologie der gattung Dercetis Agassiz (Teleostei, Pisces). Palaont. 7.., 
Stuttgart, 40 : 205-217, pis. 19-21, 6 figs. 

Signeux, J. 1954. Notes paleoichthyologiques. IV. Leptotrachelus nouveau du Senonien 
de Sahel Alma (Liban). Bull. Mus. naln. Hist, nat., Paris, (2) 26, 5 : 642-643, 1 pi. 

Toombs, H. A. 1948. The use of acetic acid in the development of vertebrate fossils. 
Museums J., London, 48 : 54-55, 1 pi. 

& Rixon, A. E. 1950. The use of plastics in the 'transfer method' of preparing fossils. 
Museums /., London, 50 : 105-107. 

Vladykov, V. I). 1962. Osteological studies on Pacific Salmon of the genus Oncorhynchus. 
Bull. Fish. Res. Bd. Can., Ottawa, 136 : 1-172, 89 figs. 

\\ i.nzMAN, S. H. 1967. The origin of the stomiatoid fishes with comments on the classifica- 
tion of salmoniform fishes. Copeia, Washington, 1967 : 507-540, 18 figs. 



ESPECIALLY M YCTOPHOIDS 



253 



White, E. I. & Moy-Thomas, J. A. 1940. Notes on the nomenclature of fossil fishes — Part II, 

Homonyms D-L. Ann. Mag. nat. Hist., London, (2) 6 : 98-103. 

1941. Ibid. Part III, Homonyms M-Z. Ann. Mag. nat. Hist., London, (2) 7 : 395-400. 

Winkler, T. C. 1876. Deuxieme memoire sur les dents de poissons fossiles du terrain 

bruxellien. Archs. Mus. Teyler, Haarlem, 4 : 16-48, 1 pi. 
Woodward, A. S. 1888a. A synopsis of the Vertebrate fossils of the English Chalk. Proc. 

Geol. Ass., London, 10 : 273-338, 1 pi., 1 fig. 

1888b. A comparison of the Cretaceous fish fauna of Mount Lebanon with that of the 

English Chalk, (abstract). Ann. Mag. nat. Hist., London, (6) 2 : 354-355. 

1891. On some Upper Cretaceous fishes of the family Aspidorhynchidae. Proc. zool. Soc. 

Lond., 1890 : 629-637, pis. 54 & 55. 
1898. Notes on some type specimens of Cretaceous fishes from Mount Lebanon in the 

Edinburgh Museum of Science and Art. Ann. Mag. nat. Hist., London, (7) 2 : 405-414. 

1901. Catalogue of the fossil fishes in the British Museum (Natural History), 4. xxxviii + 

636 pp., 19 pis., 22 figs. Brit. Mus. (Nat. Hist.) London. 

1902. The fossil fishes of the English Chalk. Part I. Palaeontogr. Soc. (Monogr.)., 

London, 1902 : 1-56, pis. 1-13. 

1903. Ibid. Part 2. Palaeontogr. Soc. (Monogr.)., London, 1903 : 57-96, pis. 14-20. 

1912. Ibid. Part 7. Palaeontogr. Soc. (Monogr.)., London, 1912 : 225-264, pis. 47-54. 

1926. Fossil fishes in the Lebanon. The Illustrated London News, 4th Sept., 1926 : 

398-9, 6 figs. 

1932. Zittel, K. A., Text-book of Palaeontology, 2. xvii + 464 pp., 533 figs. London : 

Macmillan (Revised English edition). 

1942a. Some new and little known Upper Cretaceous fishes from Mount Lebanon. Ann. 

Mag. nat. Hist., London, (11) 9 : 537-568, 4 figs., 5 pis. 

1942b. The beginning of the teleostean fishes. Ann. Mag. nat. Hist., London, (n) 

9 : 902-912, 1 fig. 

& Sherborn, C. D. 1890. A catalogue of British Fossil Vertebrata. xxxv + 396 pp. 

London : Dulau and Co. 



VIII. LETTERING USED IN THE TEXT-FIGURES 



a.e.pal 


anterior extent of the palatines 


c 


a.e.pm 


anterior extent of the 


ch 




premaxillae 


cl 


ang 


angular 


cor 


ant 


antorbital 


c.pt.f 


ar.p.pm 


articular process of the maxilla 




art 


articular 




art.f 


articular flange lateral to the 






articular facet 


den 

den t 


as.p.pm 


ascending process of the 




premaxilla 


d.f 


a.t.f.c. 


anterior opening of the pars 
jugularis 


d.pcl 
d.pt 


b.f 


basal fulcral scale 




b.md.V 


foramen of branch of mandibu- 






lar (trigeminal) nerve 


ecp 


bo 


basioccipital 


ecp.t 


brs 


branchiostegal rays 


enp 


bs 


basisphenoid 


enp.t 



centrum 

ceratohyal 

cleithrum 

coracoid 

interspace of cartilage in the 

floor of the post-temporal 

fossa 

dentary 

anterior, enlarged dentary tooth 
dilatator fossa 
dorsal postcleithrum 
dorsal limb of the post- 
temporal 

ectopterygoid 
ectopterygoid tooth 
endopterygoid 
endopterygoid teeth 



^54 



UPPER CRETACEOUS TELEOSTS 




f.art 
f.e.p.a. 

f.f 
f.hm 

f.i.c.a. 

f.l.e 
f.m 
f.mes.vo 

f.o.n 
f.o.a 
f.p 
f.d.pt 

f.v.pt 

fr 

f i. ecp \ 
i 2. ecp I 



g.md. VII 



g.sym 



h.a 
hh 
hm 
hm.h 
hm. VII 



h.mx 
h.s 

hy 

ic 
inf 

inf.s.c 
iop 



ist, 2nd and 3rd epural bones 

epiotic 
exoccipital 

articular facet 
foramen of the efferent 

pseudobranchial artery 
facial foramen 
articular facet for the 

hyomandibular 
foramen of the internal carotid 

artery 
facet for the lateral ethmoid 
foramen magnum 
facet for the mesethmoid and 

vomer 
foramen for occipital nerve 
foramen for orbital artery 
facet for palatine 
facet for the dorsal limb of the 

post-temporal 
facet for the ventral limb of the 

post-temporal 
frontal 
facets for the articulation of the 

palatine with the 

ectopterygoid 

groove for the mandibular 
branch of the facial nerve 

symplectic groove of the 
quadrate 

haemal arch 

hypohyal 

hyomandibular 

head of the hyomandibular 

foramen of the hyomandibular 
nerve within the hyomandi- 
bular bone 

articular head of the maxilla 

haemal spine 

hypural (numbered 1-6) 

intercalar 

infraorbital bone (numbered 

1-6) 
infraorbital sensory canal 
interoperculum 



la 


lachrymal 


l.e 


lateral ethmoid 


1.1 


course of the lateral line 




sensory canal 


l.l.sc 


lateral line scale 


m.cor 


mesocoracoid arch 


med.s.c 


sensory canal in the mid-dorsal 




line 


mes 


mesethmoid 


mpt 


metapterygoid 


mx 


maxilla 


mx.g 


maxillary groove on the 




palatine 


mx.p.p 


maxillary process of the 




palatine 


myo 


myodome 


n.a 


neural arch 


na 


nasal 


na.pu 1 


neural arch of the first preural 




vertebra 


na.pu 2 


neural arch of the second 




preural vertebra 


na.u 1 


neural arch of the first ural 




vertebra 


n.s 


neural spine 


0. inf.s.c 


otic branch of the infraorbital 




sensory canal 


op 


operculum 


op. VII 


foramen of the opercular 




branch of the facial nerve 


op.p.hm 


opercular process of the 




hyomandibular 


ors 


orbitosphenoid 


ot. VII 


foramen of the otic branch of 




the facial nerve 


p. VII 


course of the palatine branch of 




the facial nerve 


pa 


parietal 


pal 


palatine 


pal.t 


palatine tooth 


par 


parasphenoid 


pel 


postcleithrum 


p.e.f 


fenestra entering the cranial 




cavity below the post- 




temporal fossa 


p.f 


profundus foramen 


ph 


parhypural 


p. ecp 


palatine prominence for 




articulation with the 




ectopterygoid 



v 



ESPECIALLY MYCTOPHOIDS 



*55 



pis pleurosphenoid 

pm premaxilla 

pm.f fenestra in the ascending 

process of the premaxilla 

pop preoperculum 

pop.s preopercular spine 

post.z postzygapophysis 

pop.s.c preopercular sensory canal 

prez prezygapophysis 

pro prootic 

pro.b prootic bridge 

pt post-temporal 

pt.f post-temporal fossa 

p.t.f.c. posterior opening of the pars 

jugularis 

pto pterotic 

pu preural vertebra (numbered 1-9) 

pu 1 + u 1 preural vertebra one fused with 

ural vertebra one 

p.vo prominence for articulation 

with the vomer 

q quadrate 

r.n.s.pu 2 reduced neural spine of the 2nd 
preural vertebra 

s.c foramen of sensory canal 

sea scapula 

sea. cor scapulocoracoid 

sea. for scapular foramen 

scl supracleithrum 

smx supramaxilla 

smx 1 anterior supramaxilla 

smx 2 posterior supramaxilla 

so supraorbital 



so. s.c 

so. V, VII 

soc 
sop 
spo 
s.pop 

St 

st.s.c 

su 

sym 

t.f 
t.ps 

U I 
U 2 

ur 1 
ur 2 

vo 

vo.t 

v.pcl 

v.pt 



III 

Vll.hm 
VII. p 
IX 
X 



supraorbital sensory canal 
foramen of the superficial 

ophthalmic nerves 
supraoccipital 
suboperculum 
sphenotic 

suprapreoperculum 
supratemporal 
supratemporal sensory canal 
stegural 
symplectic 

trigeminal foramen 
transverse process 

ural vertebrae one and two 
uroneurals one and two 



vomer 

vomerine teeth 
ventral postcleithrum 
ventral limb of the post- 
temporal 

foramen of the oculomotor 

nerve 
foramen 

nerve 
foramen 

of the 
foramen 

nerve 
foramen of the vagus nerve 



of the hyomandibular 

in the prootic 

for the palatine branch 

facial nerve 

of the glossopharyngeal 




P. C. Goody, B.Sc, Ph.D., 
Royal Veterinary College, 
Royal College Street, 
London, N.W.i 



- . k 







A LIST OF SUPPLEMENTS 
TO THE GEOLOGICAL SERIES 
OF THE BULLETIN OF 
THE BRITISH MUSEUM (NATURAL HISTORY) 



i. Cox, L. R. Jurassic Bivalvia and Gastropoda from Tanganyika and Kenya. 
Pp. 213 ; 30 plates, 2 Text-figures. 1965. £6. 

2. El-Naggar, Z. R. Stratigraphy and Planktonic Foraminifera of the Upper 
Cretaceous — Lower Tertiary Succession in the Esna-Idfu Region, Nile Valley, 
Egypt, U.A.R. Pp. 291 ; 23 plates, 18 Text-figures. 1966. £10. 

3. Davey, R. J., Downie, C, Sargeant, W. A. S. & Williams, G. L. Studies on 
Mesozoic and Cainozoic Dinoflagellate Cysts. Pp. 248; 28 plates, 64 Text- 
figures. 1966. £y. 

4. Elliott, G. F. Permian to Palaeocene Calcareous Algae (Dasycladaceae) of the 
Middle East. Pp. in; 24 plates, 17 Text-figures. 1968. £5 2s. 6d. 

5. Rhodes, F. H. T., Austin, R. L. & Druce, E. C. British Avonian (Carboni- 
ferous) Conodont faunas, and their value in local and continental correlation. 
Pp- 3*5 '> 3 1 plates, 92 Text-figures. 1969. £n. 

6. Childs, A. Some Upper Jurassic Rhynchonellid Brachiopods from Northwest 
Europe. Pp. 119; 12 plates, 40 Text figures. 1969, £4. 15s. 



Printed in England by Staples Printers Limited at their Kettering, Northants, establishment